Compare commits

..

182 Commits

Author SHA1 Message Date
YoVinchen fdb36ead45 fix(usage): reorder cache columns and prevent header text wrapping
- Swap cache read and cache write columns order
- Add whitespace-nowrap to all table headers to prevent text wrapping
- Improves table readability and layout consistency
2025-12-07 14:36:57 +08:00
YoVinchen 663acf49e8 feat(skill): add multi-app skill support for Claude/Codex/Gemini
- Add app-specific skill management with AppType prefix in skill keys
- Implement per-app skill tracking in database schema
- Add get_skills_for_app command to retrieve skills by application
- Update SkillsPage to support app-specific skill loading with initialApp prop
- Parse app parameter and validate against supported app types
- Maintain backward compatibility with default claude app
2025-12-07 14:08:03 +08:00
YoVinchen 622a24ded4 fix(skill): use directory basename for skill installation path (#358)
Extract last segment from skill directory path to prevent nested directory
issues during install/uninstall operations. For example, "skills/codex" now
correctly installs to "codex" instead of creating nested "skills/codex" path.
2025-12-05 14:57:06 +08:00
Jason 6713368657 fix(windows): use system titlebar to prevent black screen on startup
The `titleBarStyle: "Overlay"` setting in tauri.conf.json causes black
screen and crash on Windows due to WebView2 compatibility issues.

Add platform-specific config `tauri.windows.conf.json` to override
titleBarStyle to "Visible" on Windows while keeping the overlay style
on macOS for the immersive UI experience.

Fixes black screen issue on Windows for v3.8.x releases.
2025-12-05 11:44:54 +08:00
YoVinchen b1103c8a59 Feat/proxy server (#355)
* feat(proxy): implement local HTTP proxy server with multi-provider failover

Add a complete HTTP proxy server implementation built on Axum framework,
enabling local API request forwarding with automatic provider failover
and load balancing capabilities.

Backend Implementation (Rust):
- Add proxy server module with 7 core components:
  * server.rs: Axum HTTP server lifecycle management (start/stop/status)
  * router.rs: API routing configuration for Claude/OpenAI/Gemini endpoints
  * handlers.rs: Request/response handling and transformation
  * forwarder.rs: Upstream forwarding logic with retry mechanism (652 lines)
  * error.rs: Comprehensive error handling and HTTP status mapping
  * types.rs: Shared types (ProxyConfig, ProxyStatus, ProxyServerInfo)
  * health.rs: Provider health check infrastructure

Service Layer:
- Add ProxyService (services/proxy.rs, 157 lines):
  * Manage proxy server lifecycle
  * Handle configuration updates
  * Track runtime status and metrics

Database Layer:
- Add proxy configuration DAO (dao/proxy.rs, 242 lines):
  * Persist proxy settings (listen address, port, timeout)
  * Store provider priority and availability flags
- Update schema with proxy_config table (schema.rs):
  * Support runtime configuration persistence

Tauri Commands:
- Add 6 command endpoints (commands/proxy.rs):
  * start_proxy_server: Launch proxy server
  * stop_proxy_server: Gracefully shutdown server
  * get_proxy_status: Query runtime status
  * get_proxy_config: Retrieve current configuration
  * update_proxy_config: Modify settings without restart
  * is_proxy_running: Check server state

Frontend Implementation (React + TypeScript):
- Add ProxyPanel component (222 lines):
  * Real-time server status display
  * Start/stop controls
  * Provider availability monitoring
- Add ProxySettingsDialog component (420 lines):
  * Configuration editor (address, port, timeout)
  * Provider priority management
  * Settings validation
- Add React hooks:
  * useProxyConfig: Manage proxy configuration state
  * useProxyStatus: Poll and display server status
- Add TypeScript types (types/proxy.ts):
  * Define ProxyConfig, ProxyStatus interfaces

Provider Integration:
- Extend Provider model with availability field (providers.rs):
  * Track provider health for failover logic
- Update ProviderCard UI to display proxy status
- Integrate proxy controls in Settings page

Dependencies:
- Add Axum 0.7 (async web framework)
- Add Tower 0.4 (middleware and service abstractions)
- Add Tower-HTTP (CORS layer)
- Add Tokio sync primitives (oneshot, RwLock)

Technical Details:
- Graceful shutdown via oneshot channel
- Shared state with Arc<RwLock<T>> for thread-safe config updates
- CORS enabled for cross-origin frontend access
- Request/response streaming support
- Automatic retry with exponential backoff (forwarder)
- API key extraction from multiple config formats (Claude/Codex/Gemini)

File Statistics:
- 41 files changed
- 3491 insertions(+), 41 deletions(-)
- Core modules: 1393 lines (server + forwarder + handlers)
- Frontend UI: 642 lines (ProxyPanel + ProxySettingsDialog)
- Database/DAO: 326 lines

This implementation provides the foundation for advanced features like:
- Multi-provider load balancing
- Automatic failover on provider errors
- Request logging and analytics
- Usage tracking and cost monitoring

* fix(proxy): resolve UI/UX issues and database constraint error

Simplify proxy control interface and fix database persistence issues:

Backend Fixes:
- Fix NOT NULL constraint error in proxy_config.created_at field
  * Use COALESCE to preserve created_at on updates
  * Ensure proper INSERT OR REPLACE behavior
- Remove redundant enabled field validation on startup
  * Auto-enable when user clicks start button
  * Persist enabled state after successful start
- Preserve enabled state during config updates
  * Prevent accidental service shutdown on config save

Frontend Improvements:
- Remove duplicate proxy enable switch from settings dialog
  * Keep only runtime toggle in ProxyPanel
  * Simplify user experience with single control point
- Hide proxy target button when proxy service is stopped
  * Add isProxyRunning prop to ProviderCard
  * Conditionally render proxy controls based on service status
- Update form schema to omit enabled field
  * Managed automatically by backend

Files: 5 changed, 81 insertions(+), 94 deletions(-)

* fix(proxy): improve URL building and Gemini request handling

- Refactor URL construction with version path deduplication (/v1, /v1beta)
- Preserve query parameters for Gemini API requests
- Support GOOGLE_GEMINI_API_KEY field name (with fallback)
- Change default proxy port from 5000 to 15721
- Fix test: use Option type for is_proxy_target field

* refactor(proxy): remove unused request handlers and routes

- Remove unused GET/DELETE request forwarding methods
- Remove count_tokens, get/delete response handlers
- Simplify router by removing unused endpoints
- Keep only essential routes: /v1/messages, /v1/responses, /v1beta/*

* Merge branch 'main' into feat/proxy-server

* fix(proxy): resolve clippy warnings for dead code and uninlined format args

- Add #[allow(dead_code)] to unused ProviderUnhealthy variant
- Inline format string arguments in handlers.rs and codex.rs log macros
- Refactor error response handling to properly pass through upstream errors
- Add URL deduplication logic for /v1/v1 paths in CodexAdapter

* feat(proxy): implement provider adapter pattern with OpenRouter support

This major refactoring introduces a modular provider adapter architecture
to support format transformation between different AI API formats.

New features:
- Add ProviderAdapter trait for unified provider abstraction
- Implement Claude, Codex, and Gemini adapters with specific logic
- Add Anthropic ↔ OpenAI format transformation for OpenRouter compatibility
- Support model mapping from provider configuration (ANTHROPIC_MODEL, etc.)
- Add OpenRouter preset to Claude provider presets

Refactoring:
- Extract authentication logic into auth.rs with AuthInfo and AuthStrategy
- Move URL building and request transformation to individual adapters
- Simplify ProviderRouter to only use proxy target providers
- Refactor RequestForwarder to use adapter-based request/response handling
- Use whitelist mode for header forwarding (only pass necessary headers)

Architecture:
- providers/adapter.rs: ProviderAdapter trait definition
- providers/auth.rs: AuthInfo, AuthStrategy types
- providers/claude.rs: Claude adapter with OpenRouter detection
- providers/codex.rs: Codex (OpenAI) adapter
- providers/gemini.rs: Gemini (Google) adapter
- providers/models/: Anthropic and OpenAI API data models
- providers/transform.rs: Bidirectional format transformation

* feat(proxy): add streaming SSE transform and thinking parameter support

New features:
- Add OpenAI → Anthropic SSE streaming response transformation
- Support thinking parameter detection for reasoning model selection
- Add ANTHROPIC_REASONING_MODEL config option for extended thinking

Changes:
- streaming.rs: Implement SSE event parsing and Anthropic format conversion
- transform.rs: Add thinking detection logic and reasoning model mapping
- handlers.rs: Integrate streaming transform for OpenRouter compatibility
- Cargo.toml: Add async-stream and bytes dependencies

* feat(db): add usage tracking schema and types

Add database tables for proxy request logs and model pricing.
Extend Provider and error types to support usage statistics.

* feat(proxy): implement usage tracking subsystem

Add request logger with automatic cost calculation.
Implement token parser for Claude/OpenAI/Gemini responses.
Add cost calculator based on model pricing configuration.

* feat(proxy): integrate usage logging into request handlers

Add usage logging to forwarder and streaming handlers.
Track token usage and costs for each proxy request.

* feat(commands): add usage statistics Tauri commands

Register usage commands for summary, trends, logs, and pricing.
Expose usage stats service through Tauri command layer.

* feat(api): add frontend usage API and query hooks

Add TypeScript types for usage statistics.
Implement usage API with Tauri invoke calls.
Add TanStack Query hooks for usage data fetching.

* feat(ui): add usage dashboard components

Add UsageDashboard with summary cards, trend chart, and data tables.
Implement model pricing configuration panel.
Add request log viewer with filtering and detail panel.

* fix(ui): integrate usage dashboard and fix type errors

Add usage dashboard tab to settings page.
Fix UsageScriptModal TypeScript type annotations.

* deps: add recharts for charts and rust_decimal/uuid for usage tracking

- recharts: Chart visualization for usage trends
- rust_decimal: Precise cost calculations
- uuid: Request ID generation

* feat(proxy): add ProviderType enum for fine-grained provider detection

Introduce ProviderType enum to distinguish between different provider
implementations (Claude, ClaudeAuth, Codex, Gemini, GeminiCli, OpenRouter).
This enables proper authentication handling and request transformation
based on the actual provider type rather than just AppType.

- Add ProviderType enum with detection logic from config
- Enhance Claude adapter with OpenRouter detection
- Enhance Gemini adapter with CLI mode detection
- Add helper methods for provider type inference

* feat(database): extend schema with streaming and timing fields

Add new columns to proxy_request_logs table for enhanced usage tracking:
- first_token_ms and duration_ms for performance metrics
- provider_type and is_streaming for request classification
- cost_multiplier for flexible pricing

Update model pricing with accurate rates for Claude/GPT/Gemini models.
Add ensure_model_pricing_seeded() call on database initialization.
Add test for model pricing auto-seeding verification.

* feat(proxy/usage): enhance token parser and logger for multi-format support

Parser enhancements:
- Add OpenAI Chat Completions format parsing (prompt_tokens/completion_tokens)
- Add model field to TokenUsage for actual model name extraction
- Add from_codex_response_adjusted() for proper cache token handling
- Add debug logging for better stream event tracing

Logger enhancements:
- Add first_token_ms, provider_type, is_streaming, cost_multiplier fields
- Extend RequestLog struct with full metadata tracking
- Update log_with_calculation() signature for new fields

Calculator: Update tests with model field in TokenUsage.

* feat(proxy): enhance proxy server with session tracking and OpenAI route

Error handling:
- Add StreamIdleTimeout and AuthError variants for better error classification

Module exports:
- Export ResponseType, StreamHandler, NonStreamHandler from response_handler
- Export ProxySession, ClientFormat from session module

Server routing:
- Add /v1/chat/completions route for OpenAI Chat Completions API

Handlers:
- Add log_usage_with_session() for enhanced usage tracking with session context
- Add first_token_ms timing measurement for streaming responses
- Use SseUsageCollector with start_time for accurate latency calculation
- Track is_streaming flag in usage logs

* feat(services): add pagination and enhanced filtering for request logs

Usage stats service:
- Change get_request_logs() from limit/offset to page/page_size pagination
- Return PaginatedLogs with total count, page, and page_size
- Add appType and providerName filters with LIKE search
- Add is_streaming, first_token_ms, duration_ms to RequestLogDetail
- Join with providers table for provider name lookup

Commands:
- Update get_request_logs command signature for pagination params

Module exports:
- Export PaginatedLogs struct

* feat(frontend): update usage types and API for pagination support

Types (usage.ts):
- Add isStreaming, firstTokenMs, durationMs to RequestLog
- Add PaginatedLogs interface with data, total, page, pageSize
- Change LogFilters: providerId -> appType + providerName

API (usage.ts):
- Change getRequestLogs params from limit/offset to page/pageSize
- Return PaginatedLogs instead of RequestLog[]
- Pass filters object directly to backend

Query (usage.ts):
- Update usageKeys.logs key generation for pagination
- Update useRequestLogs hook signature

* refactor(ui): enhance RequestLogTable with filtering and pagination

UI improvements:
- Add filter bar with app type, provider name, model, status selectors
- Add date range picker (startDate/endDate)
- Add search/reset/refresh buttons

Pagination:
- Implement proper page-based pagination with page info display
- Show total count and current page range
- Add prev/next navigation buttons

Features:
- Default to last 24 hours filter
- Streamlined table columns layout
- Query invalidation on refresh

* style(config): format mcpPresets code style

Apply consistent formatting to createNpxCommand function and
sequential-thinking server configuration.

* fix(ui): update SettingsPage tab styles for improved appearance (#342)

* feat(model-test): add provider model availability testing

Implement standalone model testing feature to verify provider API connectivity:
- Add ModelTestService for Claude/Codex/Gemini endpoint testing
- Create model_test_logs table for test result persistence
- Add test button to ProviderCard with loading state
- Include ModelTestConfigPanel for customizing test parameters

* fix(proxy): resolve token parsing for OpenRouter streaming responses

Problem:
- OpenRouter and similar third-party services return streaming responses
  where input_tokens appear in message_delta instead of message_start
- The previous implementation only extracted input_tokens from message_start,
  causing input_tokens to be recorded as 0 for these providers

Changes:
- streaming.rs: Add prompt_tokens field to Usage struct and include
  input_tokens in the transformed message_delta event when converting
  OpenAI format to Anthropic format
- parser.rs: Update from_claude_stream_events() to handle input_tokens
  from both message_start (native Claude API) and message_delta (OpenRouter)
  - Use if-let pattern instead of direct unwrap for safer parsing
  - Only update input_tokens from message_delta if not already set
- logger.rs: Adjust test parameters to match updated function signature

Tests:
- Add test_openrouter_stream_parsing() for OpenRouter format validation
- Add test_native_claude_stream_parsing() for native Claude API validation

* fix(pricing): standardize model ID format for pricing lookup

Normalize model IDs by removing vendor prefixes and converting dots to hyphens to ensure consistent pricing lookups across different API response formats.

Changes:
- Update seed data to use hyphen format (e.g., gpt-5-1, gemini-2-5-pro)
- Add normalize_model_id() function to strip vendor prefixes (anthropic/, openai/)
- Convert dots to hyphens in model IDs (claude-haiku-4.5 → claude-haiku-4-5)
- Try both original and normalized IDs for exact matching
- Use normalized ID for suffix-based fallback matching
- Add comprehensive test cases for prefix and dot handling
- Add warning log when no pricing found

This ensures pricing lookups work correctly for:
- Models with vendor prefixes: anthropic/claude-haiku-4.5
- Models with dots in version: claude-sonnet-4.5
- Models with date suffixes: claude-haiku-4-5-20240229

* style(rust): apply clippy formatting suggestions

Apply automatic clippy fixes for uninlined_format_args warnings across Rust codebase. Replace format string placeholders with inline variable syntax for improved readability.

Changes:
- Convert format!("{}", var) to format!("{var}")
- Apply to model_test.rs, parser.rs, and usage_stats.rs
- Fix line length issues by breaking long function calls
- Improve code formatting consistency

All changes are automatic formatting with no functional impact.

* fix(ui): restore card borders in usage statistics panels

Restore proper card styling for ModelTestConfigPanel and PricingConfigPanel by adding back border and rounded-lg classes. The transparent background styling was causing visual inconsistency.

Changes:
- Replace border-none bg-transparent shadow-none with border rounded-lg
- Apply to both loading and error states for consistency
- Format TypeScript code for better readability
- Break long function signatures across multiple lines

This ensures the usage statistics panels have consistent visual appearance with proper borders and rounded corners.

* feat(pricing): add GPT-5 Codex model pricing presets

Add pricing configuration for GPT-5 Codex variants to support cost tracking for Codex-specific models.

Changes:
- Add gpt-5-codex model with standard GPT-5 pricing
- Add gpt-5-1-codex model with standard GPT-5.1 pricing
- Input: $1.25/M tokens, Output: $10/M tokens
- Cache read: $0.125/M tokens, Cache creation: $0

This ensures accurate cost calculation for Codex API requests using GPT-5 Codex models.
2025-12-05 11:26:41 +08:00
Jason bf9228093b fix: add fallback for crypto.randomUUID() on older WebViews
crypto.randomUUID() is not available on older systems (macOS < 12.3,
Safari < 15.4). This caused "crypto.randomUUID is not a function"
error when adding providers.

- Add generateUUID() utility with getRandomValues() fallback
- Provide user-friendly error message if crypto API unavailable
2025-12-04 17:01:17 +08:00
Forte Scarlet 420b5d4389 feat: 为切换弹出提示框增加可关闭按钮 (#350) 2025-12-04 15:04:39 +08:00
wenyuan b1e4c37f9c fix(ui): update SettingsPage tab styles for improved appearance (#342) 2025-12-03 10:21:37 +08:00
farion1231 74a2e9c08b fix(mcp): use browser-compatible platform detection for MCP presets
Replace Node.js process.platform check with navigator.userAgent-based
detection from @/lib/platform. The previous implementation using
process.platform failed in Tauri's browser environment, causing
Windows platform to always return false and not apply the required
'cmd /c' wrapper for npx commands.

This fix ensures that on Windows, MCP presets like sequential-thinking
correctly use 'cmd /c npx' format instead of direct 'npx', eliminating
the warning: "Windows requires 'cmd /c' wrapper to execute npx"

Changes:
- Import isWindows() from @/lib/platform instead of defining locally
- Remove incorrect process.platform-based detection
- Now properly detects Windows in browser/WebView environment
2025-12-02 15:43:14 +08:00
farion1231 169db5b6d8 chore: add nul to .gitignore to prevent Windows reserved filename tracking 2025-12-02 10:00:15 +08:00
Jason 3230b0e094 chore: bump version to 3.8.2 2025-12-01 22:41:43 +08:00
Jason 2347ac0ee0 fix(config): update DeepSeek default model from Exp to stable version 2025-12-01 22:37:01 +08:00
Jason 58c5468bf6 fix(ui): prevent provider card hover scale from being clipped
Move 1px of horizontal padding from outer container to inner scroll
container, providing buffer space for the hover scale-[1.01] effect
without being cut off by overflow-hidden.
2025-12-01 16:03:20 +08:00
Jason 98084d61aa fix(ui): add independent scroll containers to fix scroll wheel on Linux
Providers page was using DndContext which may interfere with scroll wheel
events on Linux/Ubuntu WebKitGTK. Added independent scroll containers
with `overflow-y-auto` to all main pages, matching the pattern already
used by the MCP panel which works correctly.

Changes:
- App.tsx: Wrap ProviderList in independent scroll container
- SkillsPage: Use consistent h-[calc(100vh-8rem)] layout
- SettingsPage: Add overflow-hidden and overflow-x-hidden for consistency
2025-12-01 11:28:01 +08:00
Jason a627e1bb50 fix(config): update Codex default model and correct MiniMax URL
- Update Codex preset model from gpt-5-codex to gpt-5.1-codex
- Fix MiniMax English preset URL from minimaxi.io to minimax.io
2025-12-01 11:08:53 +08:00
Jason 04a588694b chore: bump version to 3.8.1 2025-11-30 23:39:05 +08:00
Jason f5f7dfed8c style: apply code formatting fixes
- Fix Prettier formatting in claudeProviderPresets.ts
- Fix cargo fmt trailing blank line in provider/mod.rs
2025-11-30 23:30:27 +08:00
Jason 5888c56f2a fix(config): correct MiniMax apiKeyUrl to minimaxi.com
Update apiKeyUrl domain to match official website URL.
Also update screenshots for all locales (en, ja, zh).
2025-11-30 23:03:09 +08:00
Jason c5cadb73bc fix(config): correct MiniMax apiKeyUrl domain from minimax.io to minimax.com 2025-11-29 23:03:19 +08:00
Jason 17948ee031 docs: update screenshots and add Japanese locale images
- Update README_JA.md to reference Japanese screenshots
- Add Japanese screenshots (main-ja.png, add-ja.png)
- Update English and Chinese screenshots with latest UI
- Remove outdated README_i18n.md documentation
2025-11-29 22:14:11 +08:00
Jason 2643595012 fix(skills): target first span only when hiding select checkmark indicator 2025-11-29 21:36:16 +08:00
Jason 6ac4d1652c fix(skills): use theme-aware glass-card class for light mode compatibility
Replace hardcoded dark mode styles (bg-gray-900/40, border-white/10) with
the unified glass-card class that adapts to both light and dark themes.
2025-11-29 21:18:51 +08:00
Jason 3bf37cf0ff refactor(url): remove automatic trailing slash stripping from base URL inputs
- Remove `.replace(/\/+$/, "")` from all base URL handlers in useBaseUrlState.ts
- Remove trailing slash stripping in useCodexConfigState.ts
- Remove trailing slash normalization in providerConfigUtils.ts setCodexBaseUrl()
- Update i18n hints (en/ja/zh) to instruct users to avoid trailing slashes

This gives users explicit control over URL format rather than silently modifying input.
2025-11-29 20:52:29 +08:00
Jason 526c7406fa fix(ui): persist active provider card highlight when not hovered
The selected provider's visual highlight (background, border, glow)
was being overridden by .glass-card base styles due to CSS specificity.
Add dedicated .glass-card-active class to ensure active state persists.
2025-11-29 20:19:17 +08:00
Jason 9db85dd4dc style(css): consolidate global selector rules and remove duplicates
- Merge three separate `*` selector blocks into one for better maintainability
- Remove duplicate "Glassmorphism Utilities" comment
2025-11-29 20:03:19 +08:00
Jason 0f333d9e5b fix(tailwind): add missing shadcn/ui color mappings for opaque dialog backgrounds
The dialog backgrounds were transparent because Tailwind 3 requires explicit
color mappings in the config to use CSS variables. Added standard shadcn/ui
color mappings (background, foreground, card, primary, etc.) and removed
unnecessary overlayClassName overrides from dialog components.
2025-11-29 19:49:55 +08:00
Jason ba875552a6 style(mcp): align wizard modal with confirm dialog styling
- Use zIndex="alert" and semi-transparent overlay for consistency
- Apply border-b-0/border-t-0 bg-transparent to header/footer
- Replace emerald accent with blue to match app theme
- Use Input component instead of raw input elements
- Simplify button variants (outline for cancel)
2025-11-29 17:41:12 +08:00
Jason 75e7f9d731 chore: remove dead code from MCP and provider modules
- Remove unused `read_mcp_json` from gemini_mcp.rs
- Remove unused `normalize_server_keys` from mcp/claude.rs
- Remove unused `validate_mcp_entry` from mcp/validation.rs
- Remove unused `write_codex_live`, `write_claude_live`, `app_not_found` from services/provider/mod.rs
- Clean up unused imports
2025-11-29 16:24:22 +08:00
Jason 7e6074a9a9 refactor(mcp): modularize MCP and tray menu logic
Split mcp.rs (1135 lines) into focused modules:
- validation.rs: server config validation
- claude.rs: Claude MCP sync/import
- codex.rs: Codex MCP sync (TOML handling)
- gemini.rs: Gemini MCP sync/import

Extract tray menu logic from lib.rs to tray.rs for better separation of concerns.
2025-11-29 10:29:10 +08:00
Jason c229c47c00 fix(auto-launch): use AutoLaunchBuilder for cross-platform compatibility
The auto-launch crate has different API signatures on each platform:
- Windows/Linux: AutoLaunch::new() takes 3 arguments
- macOS: AutoLaunch::new() takes 4 arguments (includes hidden param)

The previous code used #[cfg(not(target_os = "windows"))] which incorrectly
applied macOS's 4-argument signature to Linux, causing build failures.

Switch to AutoLaunchBuilder which handles platform differences internally.
2025-11-28 22:59:09 +08:00
Jason 6c477a60f9 chore: bump version to v3.8.0 2025-11-28 22:37:32 +08:00
Jason 7eac809689 feat(preset): add MiniMax international version and split promotions
- Add MiniMax en preset with international API endpoint (minimaxi.io)
- Split MiniMax partner promotion into CN and EN versions
- Remove unnecessary icon config from Aliyun and Alibaba Lingma presets
2025-11-28 22:27:23 +08:00
Jason dafa77897b docs: add v3.8.0 release documentation and Japanese README
- Add CHANGELOG entry for v3.8.0
- Update README.md and README_ZH.md with v3.8.0 features
- Add Japanese README (README_JA.md)
- Add release notes in English, Chinese, and Japanese
2025-11-28 21:47:05 +08:00
YoVinchen 0f959112b1 refactor(skill): remove skillsPath configuration (#310)
Remove the skillsPath field from SkillRepo and Skill structs since
recursive scanning now automatically discovers skills in all directories.
Simplify the UI by removing the path input field.
2025-11-28 16:26:28 +08:00
Jason f4c284f86c fix(test): update component tests to match current implementation
- Add JsonEditor mock to McpFormModal tests (component uses CodeMirror
  instead of Textarea)
- Fix assertion for missing command error message
- Update ImportExportSection tests for new button behavior and file
  display format
2025-11-28 16:20:18 +08:00
Jason 3878a16c4f fix: pre-release fixes for code formatting, i18n, and test suite
- Run cargo fmt to fix Rust code formatting (lib.rs)
- Add missing i18n keys: migration.success, agents.title (zh/en/ja)
- Replace hardcoded strings "Agents" and "MCP" with t() calls in App.tsx
- Fix test mocks and assertions:
  - Add providersApi.updateTrayMenu to useSettings.test.tsx mock
  - Update SettingsPage mock path in App.test.tsx
  - Fix toast message assertion in integration/SettingsDialog.test.tsx
  - Add autoSaveSettings to SettingsDialog component test mock
  - Fix loading state test to check spinner instead of title
  - Update import button name matching for selected file state
  - Fix save button test to switch to advanced tab first
  - Remove obsolete cancel button test (button no longer exists)

Test results improved from 99 passed / 17 failed to 104 passed / 11 failed
2025-11-28 15:53:25 +08:00
Jason 00f78e4546 feat(i18n): add Japanese language support
- Add complete Japanese translation file (ja.json)
- Update frontend types and hooks to support "ja" language option
- Add Japanese tray menu texts in Rust backend
- Add Japanese option to language settings component
- Update Zod schema to include Japanese language validation
- Add test case for Japanese language preference
- Update i18n documentation to reflect three-language support
2025-11-28 15:14:39 +08:00
Jason 1ee1e9cb2e refactor(startup): improve first-launch import logic with per-table detection
- Replace global `is_empty_for_first_import()` with independent checks:
  - `is_mcp_table_empty()` for MCP server imports
  - `is_prompts_table_empty()` for prompt imports
  - Skills and providers already have built-in idempotency checks

- Fix misleading logs in provider import:
  - Change `import_default_config` return type from `Result<()>` to `Result<bool>`
  - Return `true` when actually imported, `false` when skipped
  - Only log success message when import actually occurred

- Add idempotency protection to `import_from_file_on_first_launch`

This allows each data type to be independently recovered if deleted,
rather than requiring all tables to be empty for any import to trigger.
2025-11-28 12:05:29 +08:00
YoVinchen 7db4b8d976 feat(skill): implement recursive scanning for skill repositories (#309)
Add recursive directory scanning to discover SKILL.md files in nested
directories. When a SKILL.md is found, treat sibling directories as
functional folders rather than separate skills.
2025-11-28 12:01:20 +08:00
Jason 924ad44e6c fix(usage): correct selectedTemplate initialization logic
Previously, selectedTemplate was initialized to null when no NEW_API
credentials were detected, which caused the credentials config section
to be hidden even for new configurations or GENERAL template users.

Now properly detects:
- NEW_API template (has accessToken or userId)
- GENERAL template (has apiKey or baseUrl)
- Default to GENERAL for new configurations (matches default code template)
2025-11-28 11:10:22 +08:00
Jason eecd6a3a2b refactor(usage): simplify usage script modal UI
- Remove redundant "Request Configuration" section (URL, method, headers, body editors)
- Move timeout and auto query interval fields to preset template section
- Simplify "Enable usage query" toggle styling
- Remove redundant hints and variables description
- Add i18n support for Base URL label
- Include "0 to disable" hint directly in auto interval label
2025-11-28 11:09:34 +08:00
farion1231 2a6980417b fix(auto-launch): use platform-specific API for AutoLaunch::new
Windows version of auto-launch crate takes 3 arguments while
Linux/macOS takes 4 (includes hidden parameter). Use conditional
compilation to handle the difference.
2025-11-28 09:58:47 +08:00
Jason 6a9a0a7a7e feat(preset): add icon configuration to provider presets
- Add icon/iconColor fields to CodexProviderPreset and GeminiProviderPreset interfaces
- Configure icons for Claude presets (DeepSeek, Zhipu, Qwen, Kimi, MiniMax, DouBao, etc.)
- Configure icons for Codex presets (OpenAI, Azure, PackyCode)
- Configure icons for Gemini presets (Google, PackyCode)
- Fix handlePresetChange to pass icon fields when resetting form

This ensures preset icons are displayed in the icon picker when selecting a provider preset.
2025-11-27 23:28:11 +08:00
Jason fcb090dd15 fix(css): downgrade Tailwind CSS from v4 to v3.4 for better browser compatibility
Tailwind CSS v4 requires modern CSS features (@layer, @property, color-mix)
that are not supported in older WebView2 versions, causing styles to fail
on some Windows 11 systems.

Changes:
- Replace @tailwindcss/vite with postcss + autoprefixer
- Downgrade tailwindcss from 4.1.13 to 3.4.17
- Convert CSS imports from v4 syntax to v3 @tailwind directives
- Add darkMode: "selector" to tailwind.config.js
- Create postcss.config.js for PostCSS pipeline

This improves compatibility with older browsers and WebView2 runtimes
while maintaining the same visual appearance.
2025-11-27 22:55:14 +08:00
Jason ce4f0c02cb fix(linux): resolve WebKitGTK DMA-BUF rendering issue and preserve user .desktop customizations
- Set WEBKIT_DISABLE_DMABUF_RENDERER=1 at startup to fix blank screen on Debian 13.2 and Nvidia GPUs
- Only register deep-link handler on first run to avoid overwriting user customizations
- Use correct Tauri path API (app.path().data_dir()) instead of dirs::data_dir() to match plugin's actual .desktop file location
2025-11-27 19:37:18 +08:00
Jason 4ca4ad6bca feat(migration): add error dialog for config.json loading failure
- Show system dialog when config.json fails to load during migration
- User can retry loading or exit the program
- Exit before database creation ensures clean retry on next launch
- Support Chinese/English localization based on system locale
2025-11-27 16:03:50 +08:00
Jason 964ead5d0d fix(provider): validate current provider ID existence before use
Add get_effective_current_provider() to validate local settings ID
against database, with automatic cleanup and fallback to DB is_current.

This fixes edge cases in multi-device cloud sync scenarios where local
settings may contain stale provider IDs:

- current(): now returns validated effective provider ID
- update(): correctly syncs live config when local ID differs from DB
- delete(): checks both local settings and DB to prevent deletion
- switch(): backfill logic now targets valid provider
- sync_current_to_live(): uses validated ID with auto-fallback
- tray menu: displays correct checkmark on startup

Also fixes test issues:
- Add missing test setup calls (mutex, reset_test_fs, ensure_test_home)
- Correct Gemini security settings path to ~/.gemini/settings.json
2025-11-27 15:01:24 +08:00
Jason 2c90ae3509 refactor(provider): use local settings for current provider selection
Complete the device-level settings separation for cloud sync support.

Backend changes:
- Modify switch() to update both local settings and database is_current
- Modify current() to read from local settings first, fallback to database
- Rename sync_current_from_db() to sync_current_to_live()
- Update tray menu to read current provider from local settings

Frontend changes:
- Update Settings interface: remove legacy fields (customEndpoints*, security)
- Add currentProviderClaude/Codex/Gemini fields
- Update settings schema accordingly

Test fixes:
- Update Gemini security tests to check ~/.gemini/settings.json
  instead of ~/.cc-switch/settings.json (security field was never
  stored in CC Switch settings)

This ensures each device maintains its own current provider selection
independently when database is synced across devices.
2025-11-27 11:42:19 +08:00
Jason ea7169abc0 refactor(settings): move device-level settings to local file storage
Separate device-level settings from database to support cloud sync scenarios
where multiple devices share the same database but need independent settings.

Changes:
- Remove database storage logic (bind_db, save_to_db, load_from_db)
- Remove legacy custom_endpoints_claude/codex fields (actual data in provider.meta)
- Add current_provider_claude/codex/gemini fields for device-level provider selection
- Add get_current_provider() and set_current_provider() helper functions
- Settings now stored only in ~/.cc-switch/settings.json

This ensures that when database is synced across devices, each device
maintains its own current provider selection independently.
2025-11-27 11:16:43 +08:00
Jason 120ac92e77 refactor(gemini): improve config handling and code consistency
- Fix envObjToString to preserve custom environment variables
- Add bidirectional MCP format conversion (httpUrl <-> url)
- Merge provider config with existing settings.json instead of overwriting
- Remove redundant is_packycode_gemini function
- Simplify is_google_official_gemini using detect_gemini_auth_type
- Add handleGeminiModelChange for consistent field handling
2025-11-27 10:21:01 +08:00
Jason a1e7961af3 fix(gemini): correctly write custom provider env to .env file
write_live_snapshot was incorrectly passing the already-extracted env
sub-field to json_to_env, which expects the full settings_config object.
This caused json_to_env to look for env.env (nested), returning an empty
HashMap and writing an empty .env file.

Fix by delegating to write_gemini_live which correctly handles env file
writing and security flag configuration in one place.
2025-11-27 09:48:37 +08:00
Jason dc79e3a3da fix(gemini): write security auth config only to Gemini settings file
- Remove redundant security.auth.selectedType from CC Switch settings
- Fix Generic provider type not writing security flag on switch
- All non-Google Official providers now correctly write "gemini-api-key"
- Delete unused ensure_packycode_security_flag function
- Clean up SecuritySettings and SecurityAuthSettings types from AppSettings
2025-11-27 09:31:54 +08:00
Jason 7adc38eb53 feat(preset): add MiniMax as official partner with Black Friday promotion
- Mark MiniMax as partner with custom theme color (#f64551)
- Add Black Friday promotional text for Starter plan ($2/mo, 80% OFF)
- Update apiKeyUrl to MiniMax coding plan subscription page
2025-11-27 08:51:54 +08:00
Jason 3f28648931 fix(query): prevent redundant usage queries when switching apps
Add staleTime and gcTime to useUsageQuery to avoid triggering
unnecessary API calls when switching between app tabs. The staleTime
is set dynamically based on the auto-refresh interval (or 5 minutes
by default), and gcTime is set to 10 minutes to preserve cache after
component unmount.
2025-11-26 19:34:00 +08:00
Jason 90d530fa7a fix(ui): improve ConfirmDialog styling for better visual consistency
- Remove header/footer borders and backgrounds for compact alert style
- Use theme-aware overlay color (bg-background/80) instead of black
- Add zIndex="alert" to ensure dialog appears above other dialogs
2025-11-26 15:36:53 +08:00
Jason 5cd2e9fb88 refactor(provider): unify validation errors to use toast notifications
Change template parameter validation from form.setError to toast.error
for consistent UX across all required field validations.
2025-11-26 12:35:33 +08:00
Jason 8bf6ce2c25 feat(provider): add required field validation with toast notifications
- Add validation for provider name (required for all providers)
- Add validation for API endpoint and API Key (required for non-official providers)
- Use toast notifications instead of inline form errors for better visibility
- Move name validation from zod schema to handleSubmit for consistent UX
- Add i18n keys: endpointRequired, apiKeyRequired
2025-11-26 12:26:02 +08:00
Jason c41f3dcccb fix(preset): use ANTHROPIC_AUTH_TOKEN for DMXAPI and correct endpoint candidates
- Change DMXAPI preset from ANTHROPIC_API_KEY to ANTHROPIC_AUTH_TOKEN
- Fix endpointCandidates that were incorrectly copied from AiHubMix
2025-11-26 11:57:47 +08:00
Jason 1caa240d6c refactor(config): remove dead code from JSON-era import/export
Remove obsolete methods that were superseded by SQLite-based
import/export in v3.7.0:

- export_config_to_path: replaced by Database::export_sql
- load_config_for_import: no longer used
- import_config_from_path: stub that only returned error

Also remove unused AppState import.
2025-11-26 11:18:23 +08:00
Jason 7ffd3ba165 fix(provider): preserve custom endpoints when updating provider
- Replace INSERT OR REPLACE with UPDATE for existing providers to avoid
  triggering ON DELETE CASCADE on provider_endpoints table
- Fix endpoint merging logic to correctly mark database endpoints as
  isCustom when they overlap with preset endpoints

The root cause was that INSERT OR REPLACE performs DELETE + INSERT under
the hood, which triggered the foreign key cascade and deleted all
associated endpoints from provider_endpoints table.
2025-11-26 10:27:07 +08:00
YoVinchen ad131486d2 fix(skill): use full key for deduplication to allow same-name skills from different repos (#299) 2025-11-26 09:39:05 +08:00
Jason 15c6e3aec8 refactor(ui): improve header toolbar layout and transitions
- Reorder toolbar buttons to keep Prompts and MCP icons aligned right
  (consistent position between Claude and Codex apps)
- Add smooth fade-in/out transition for Skills button with opacity,
  width, scale, and padding animations
- Hide Agents button temporarily (feature in development)
- Remove two divider lines for cleaner appearance
2025-11-25 23:36:48 +08:00
YoVinchen 6783a8a183 fix(provider): include icon fields when duplicating provider (#297) 2025-11-25 23:18:54 +08:00
Jason c1c85b020d fix(ui): disable overscroll bounce effect on main view
Prevents the top border line from being pulled down when scrolling
past the top of the page by setting overscroll-behavior: none on html.
2025-11-25 16:25:08 +08:00
Jason 34dad04fb6 fix(deeplink): remove unused re-exports McpImportError and McpImportResult 2025-11-25 16:19:07 +08:00
Jason 2526dbc58f feat(migration): enable silent JSON to SQLite migration with toast notification
- Remove environment variable feature gate (CC_SWITCH_ENABLE_JSON_DB_MIGRATION)
- Automatically migrate config.json to SQLite when db doesn't exist
- Archive migrated config.json as config.json.migrated for recovery
- Add migration success flag in init_status.rs (one-time consumption)
- Add get_migration_result command for frontend to query
- Show toast notification on successful migration in App.tsx
2025-11-25 16:07:54 +08:00
Jason 014a7c0e30 refactor(services): split monolithic provider.rs into modular structure
Split the 1446-line services/provider.rs into 5 focused modules:

- gemini_auth.rs (250 lines): Gemini authentication type detection
  - PackyCode, Google OAuth, and generic provider detection
  - Security flag management for different auth types

- live.rs (300 lines): Live configuration operations
  - LiveSnapshot backup/restore
  - Reading and writing live config files
  - Sync current provider to live config

- usage.rs (150 lines): Usage script execution
  - Query and test usage scripts
  - Format usage results
  - Validate usage script configuration

- endpoints.rs (80 lines): Custom endpoints management
  - CRUD operations for provider custom endpoints
  - Last-used timestamp tracking

- mod.rs (650 lines): Core provider CRUD and service facade
  - ProviderService struct with all public methods
  - Provider add/update/delete/switch operations
  - Claude model key normalization
  - Credential extraction and validation

All 107 tests pass. This improves maintainability by:
- Separating concerns into cohesive modules
- Making Gemini-specific logic easier to find and modify
- Reducing cognitive load when working on specific features
2025-11-25 12:19:26 +08:00
Jason 87190b7af3 refactor(deeplink): split monolithic deeplink.rs into modular structure
Split the 1691-line deeplink.rs into a well-organized module directory:

- mod.rs (120 lines): DeepLinkImportRequest struct and public exports
- parser.rs (321 lines): URL parsing logic for all resource types
- provider.rs (510 lines): Provider import and config merge logic
- mcp.rs (191 lines): MCP server batch import
- prompt.rs (86 lines): Prompt import
- skill.rs (52 lines): Skill repository import
- utils.rs (99 lines): Shared utilities (URL validation, Base64 decoding)
- tests.rs (384 lines): All unit tests

Benefits:
- Max file size reduced from 1691 to 510 lines
- Each resource type has its own import module
- Shared utilities extracted for reuse
- All 17 deeplink tests + 107 total tests passing
2025-11-25 12:05:33 +08:00
Jason c3f2f0798d refactor(database): split monolithic database.rs into modular structure
Split the 1788-line database.rs into a well-organized module directory:

- mod.rs (142 lines): Database struct and initialization
- schema.rs (341 lines): Table creation and schema migrations
- backup.rs (324 lines): SQL import/export and snapshot backup
- migration.rs (240 lines): JSON to SQLite data migration
- tests.rs (280 lines): Unit tests
- dao/providers.rs (254 lines): Provider CRUD operations
- dao/mcp.rs (97 lines): MCP server CRUD operations
- dao/prompts.rs (88 lines): Prompt CRUD operations
- dao/skills.rs (124 lines): Skills CRUD operations
- dao/settings.rs (65 lines): Settings key-value storage

Benefits:
- Max file size reduced from 1788 to 341 lines
- Clear separation of concerns (schema, backup, migration, DAOs)
- Easier to navigate and maintain
- All 107 tests passing with zero API changes
2025-11-25 11:56:08 +08:00
Jason 2a842e3b33 fix(provider): add backfill and Gemini security flags to switch function
The switch function was missing two important features after the SQLite
migration:

1. Backfill mechanism: Before switching providers, read the current live
   config and save it back to the current provider. This preserves any
   manual edits users made to the live config file.

2. Gemini security flags: When switching to a Gemini provider, set the
   appropriate security.auth.selectedType:
   - PackyCode providers: "gemini-api-key"
   - Google OAuth providers: "oauth-personal"

Also update tests to:
- Use the new unified MCP structure (mcp.servers) instead of the legacy
  per-app structure (mcp.codex.servers)
- Expect backfill behavior (was incorrectly marked as "no backfill")
- Remove assertions for provider-specific file deletion (v3.7.0+ uses
  SSOT, no longer creates per-provider config files)
2025-11-25 11:16:50 +08:00
Jason 1c87c8d253 Merge feat/sqlite-migration: add database schema migration system
This merge brings the SQLite migration system from feat/sqlite-migration branch:

## New Features
- Schema version control with SCHEMA_VERSION constant
- Automatic migration of missing columns for providers table
- Dry-run validation mode for schema compatibility checks
- JSON→SQLite migration feature gate (CC_SWITCH_ENABLE_JSON_DB_MIGRATION)
- Settings reload mechanism after imports

## Test Updates
- Updated tests to use SQLite database instead of config.json
- Removed obsolete import_config_from_path tests (replaced by db.import_sql)
- Fixed MCP tests to use unified McpServer structure (v3.7.0+)
- Updated provider switch tests to reflect no-backfill behavior
- Adjusted error type matching for new error variants
2025-11-25 10:33:54 +08:00
Jason fd46dde3fb fix(i18n): add missing translation keys for deeplink confirmation components
- Add i18n support to McpConfirmation, PromptConfirmation, SkillConfirmation
- Add nested translation keys: deeplink.mcp.*, deeplink.prompt.*, deeplink.skill.*
- Replace all hardcoded Chinese strings with t() function calls
- Ensure consistent localization across all deeplink import dialogs
2025-11-25 09:52:19 +08:00
Jason 60fa9654e1 fix(i18n): replace hardcoded Chinese strings with translation keys in DeepLinkImportDialog
- Add missing i18n keys for prompt, MCP, and skill import dialogs
- Replace hardcoded toast messages with t() function calls
- Add translation keys: importPrompt, importMcp, importSkill, etc.
- Ensure all user-facing text supports both zh and en locales
2025-11-25 09:41:02 +08:00
YoVinchen 2ce8a8273c Refactor/storage (#286)
* feat(components): add reusable full-screen panel components

Add new full-screen panel components to support the UI refactoring:

- FullScreenPanel: Reusable full-screen layout component with header,
  content area, and optional footer. Provides consistent layout for
  settings, prompts, and other full-screen views.

- PromptFormPanel: Dedicated panel for creating and editing prompts
  with markdown preview support. Features real-time validation and
  integrated save/cancel actions.

- AgentsPanel: Panel component for managing agent configurations.
  Provides a consistent interface for agent CRUD operations.

- RepoManagerPanel: Full-featured repository manager panel for Skills.
  Supports repository listing, addition, deletion, and configuration
  management with integrated validation.

These components establish the foundation for the upcoming settings
page migration from dialog-based to full-screen layout.

* refactor(settings): migrate from dialog to full-screen page layout

Complete migration of settings from modal dialog to dedicated full-screen
page, improving UX and providing more space for configuration options.

Changes:
- Remove SettingsDialog component (legacy modal-based interface)
- Add SettingsPage component with full-screen layout using FullScreenPanel
- Refactor App.tsx routing to support dedicated settings page
  * Add settings route handler
  * Update navigation logic from dialog-based to page-based
  * Integrate with existing app switcher and provider management
- Update ImportExportSection to work with new page layout
  * Improve spacing and layout for better readability
  * Enhanced error handling and user feedback
  * Better integration with page-level actions
- Enhance useSettings hook to support page-based workflow
  * Add navigation state management
  * Improve settings persistence logic
  * Better error boundary handling

Benefits:
- More intuitive navigation with dedicated settings page
- Better use of screen space for complex configurations
- Improved accessibility with clearer visual hierarchy
- Consistent with modern desktop application patterns
- Easier to extend with new settings sections

This change is part of the larger UI refactoring initiative to modernize
the application interface and improve user experience.

* refactor(forms): simplify and modernize form components

Comprehensive refactoring of form components to reduce complexity,
improve maintainability, and enhance user experience.

Provider Forms:
- CodexCommonConfigModal & CodexConfigSections
  * Simplified state management with reduced boilerplate
  * Improved field validation and error handling
  * Better layout with consistent spacing
  * Enhanced model selection with visual indicators
- GeminiCommonConfigModal & GeminiConfigSections
  * Streamlined authentication flow (OAuth vs API Key)
  * Cleaner form layout with better grouping
  * Improved validation feedback
  * Better integration with parent components
- CommonConfigEditor
  * Reduced from 178 to 68 lines (-62% complexity)
  * Extracted reusable form patterns
  * Improved JSON editing with syntax validation
  * Better error messages and recovery options
- EndpointSpeedTest
  * Complete rewrite for better UX
  * Real-time testing progress indicators
  * Enhanced error handling with retry logic
  * Visual feedback for test results (color-coded latency)

MCP & Prompts:
- McpFormModal
  * Simplified from 581 to ~360 lines
  * Better stdio/http server type handling
  * Improved form validation
  * Enhanced multi-app selection (Claude/Codex/Gemini)
- PromptPanel
  * Cleaner integration with PromptFormPanel
  * Improved list/grid view switching
  * Better state management for editing workflows
  * Enhanced delete confirmation with safety checks

Code Quality Improvements:
- Reduced total lines by ~251 lines (-24% code reduction)
- Eliminated duplicate validation logic
- Improved TypeScript type safety
- Better component composition and separation of concerns
- Enhanced accessibility with proper ARIA labels

These changes make forms more intuitive, responsive, and easier to
maintain while reducing bundle size and improving runtime performance.

* style(ui): modernize component layouts and visual design

Update UI components with improved layouts, visual hierarchy, and
modern design patterns for better user experience.

Navigation & Brand Components:
- AppSwitcher
  * Enhanced visual design with better spacing
  * Improved active state indicators
  * Smoother transitions and hover effects
  * Better mobile responsiveness
- BrandIcons
  * Optimized icon rendering performance
  * Added support for more provider icons
  * Improved SVG handling and fallbacks
  * Better scaling across different screen sizes

Editor Components:
- JsonEditor
  * Enhanced syntax highlighting
  * Better error visualization
  * Improved code formatting options
  * Added line numbers and code folding support
- UsageScriptModal
  * Complete layout overhaul (1239 lines refactored)
  * Better script editor integration
  * Improved template selection UI
  * Enhanced preview and testing panels
  * Better error feedback and validation

Provider Components:
- ProviderCard
  * Redesigned card layout with modern aesthetics
  * Better information density and readability
  * Improved action buttons placement
  * Enhanced status indicators (active/inactive)
- ProviderList
  * Better grid/list view layouts
  * Improved drag-and-drop visual feedback
  * Enhanced sorting indicators
- ProviderActions
  * Streamlined action menu
  * Better icon consistency
  * Improved tooltips and accessibility

Usage & Footer:
- UsageFooter
  * Redesigned footer layout
  * Better quota visualization
  * Improved refresh controls
  * Enhanced error states

Design System Updates:
- dialog.tsx (shadcn/ui component)
  * Updated to latest design tokens
  * Better overlay animations
  * Improved focus management
- index.css
  * Added 65 lines of global utility classes
  * New animation keyframes
  * Enhanced color variables for dark mode
  * Improved typography scale
- tailwind.config.js
  * Extended theme with new design tokens
  * Added custom animations and transitions
  * New spacing and sizing utilities
  * Enhanced color palette

Visual Improvements:
- Consistent border radius across components
- Unified shadow system for depth perception
- Better color contrast for accessibility (WCAG AA)
- Smoother animations and transitions
- Improved dark mode support

These changes create a more polished, modern interface while
maintaining consistency with the application's design language.

* chore: update dialogs, i18n and improve component integration

Various functional updates and improvements across provider dialogs,
MCP panel, skills page, and internationalization.

Provider Dialogs:
- AddProviderDialog
  * Simplified form state management
  * Improved preset selection workflow
  * Better validation error messages
  * Enhanced template variable handling
- EditProviderDialog
  * Streamlined edit flow with better state synchronization
  * Improved handling of live config backfilling
  * Better error recovery for failed updates
  * Enhanced integration with parent components

MCP & Skills:
- UnifiedMcpPanel
  * Reduced complexity from 140+ to ~95 lines
  * Improved multi-app server management
  * Better server type detection (stdio/http)
  * Enhanced server status indicators
  * Cleaner integration with MCP form modal
- SkillsPage
  * Simplified navigation and state management
  * Better integration with RepoManagerPanel
  * Improved error handling for repository operations
  * Enhanced loading states
- SkillCard
  * Minor layout adjustments
  * Better action button placement

Environment & Configuration:
- EnvWarningBanner
  * Improved conflict detection messages
  * Better visual hierarchy for warnings
  * Enhanced dismissal behavior
- tauri.conf.json
  * Updated build configuration
  * Added new window management options

Internationalization:
- en.json & zh.json
  * Added 17 new translation keys for new features
  * Updated existing keys for better clarity
  * Added translations for new settings page
  * Improved consistency across UI text

Code Cleanup:
- mutations.ts
  * Removed 14 lines of unused mutation definitions
  * Cleaned up deprecated query invalidation logic
  * Better type safety for mutation parameters

Overall Impact:
- Reduced total lines by 51 (-10% in affected files)
- Improved component integration and data flow
- Better error handling and user feedback
- Enhanced i18n coverage for new features

These changes improve the overall polish and integration of various
components while removing technical debt and unused code.

* feat(backend): add auto-launch functionality

Implement system auto-launch feature to allow CC-Switch to start
automatically on system boot, improving user convenience.

Backend Implementation:
- auto_launch.rs: New module for auto-launch management
  * Cross-platform support using auto-launch crate
  * Enable/disable auto-launch with system integration
  * Proper error handling for permission issues
  * Platform-specific implementations (macOS/Windows/Linux)

Command Layer:
- Add get_auto_launch command to check current status
- Add set_auto_launch command to toggle auto-start
- Integrate commands with settings API

Settings Integration:
- Extend Settings struct with auto_launch field
- Persist auto-launch preference in settings store
- Automatic state synchronization on app startup

Dependencies:
- Add auto-launch ^0.5.0 to Cargo.toml
- Update Cargo.lock with new dependency tree

Technical Details:
- Uses platform-specific auto-launch mechanisms:
  * macOS: Login Items via LaunchServices
  * Windows: Registry Run key
  * Linux: XDG autostart desktop files
- Handles edge cases like permission denials gracefully
- Maintains settings consistency across app restarts

This feature enables users to have CC-Switch readily available
after system boot without manual intervention, particularly useful
for users who frequently switch between API providers.

* refactor(settings): enhance settings page with auto-launch integration

Complete refactoring of settings page architecture to integrate auto-launch
feature and improve overall settings management workflow.

SettingsPage Component:
- Integrate auto-launch toggle with WindowSettings section
- Improve layout and spacing for better visual hierarchy
- Enhanced error handling for settings operations
- Better loading states during settings updates
- Improved accessibility with proper ARIA labels

WindowSettings Component:
- Add auto-launch switch with real-time status
- Integrate with backend auto-launch commands
- Proper error feedback for permission issues
- Visual indicators for current auto-launch state
- Tooltip guidance for auto-launch functionality

useSettings Hook (Major Refactoring):
- Complete rewrite reducing complexity by ~30%
- Better separation of concerns with dedicated handlers
- Improved state management using React Query
- Enhanced auto-launch state synchronization
  * Fetch auto-launch status on mount
  * Real-time updates on toggle
  * Proper error recovery
- Optimized re-renders with better memoization
- Cleaner API for component integration
- Better TypeScript type safety

Settings API:
- Add getAutoLaunch() method
- Add setAutoLaunch(enabled: boolean) method
- Type-safe Tauri command invocations
- Proper error propagation to UI layer

Architecture Improvements:
- Reduced hook complexity from 197 to ~140 effective lines
- Eliminated redundant state management logic
- Better error boundaries and fallback handling
- Improved testability with clearer separation

User Experience Enhancements:
- Instant visual feedback on auto-launch toggle
- Clear error messages for permission issues
- Loading indicators during async operations
- Consistent behavior across all platforms

This refactoring provides a solid foundation for future settings
additions while maintaining code quality and user experience.

* refactor(ui): optimize FullScreenPanel, Dialog and App routing

Comprehensive refactoring of core UI components to improve code quality,
maintainability, and user experience.

FullScreenPanel Component:
- Enhanced props interface with better TypeScript types
- Improved layout flexibility with customizable padding
- Better header/footer composition patterns
- Enhanced scroll behavior for long content
- Added support for custom actions in header
- Improved responsive design for different screen sizes
- Better integration with parent components
- Cleaner prop drilling with context where appropriate

Dialog Component (shadcn/ui):
- Updated to latest component patterns
- Improved animation timing and easing
- Better focus trap management
- Enhanced overlay styling with backdrop blur
- Improved accessibility (ARIA labels, keyboard navigation)
- Better close button positioning and styling
- Enhanced mobile responsiveness
- Cleaner composition with DialogHeader/Footer

App Component Routing:
- Refactored routing logic for better clarity
- Improved state management for navigation
- Better integration with settings page
- Enhanced error boundary handling
- Cleaner separation of layout concerns
- Improved provider context propagation
- Better handling of deep links
- Optimized re-renders with React.memo where appropriate

Code Quality Improvements:
- Reduced prop drilling with better component composition
- Improved TypeScript type safety
- Better separation of concerns
- Enhanced code readability with clearer naming
- Eliminated redundant logic

Performance Optimizations:
- Reduced unnecessary re-renders
- Better memoization of callbacks
- Optimized component tree structure
- Improved event handler efficiency

User Experience:
- Smoother transitions and animations
- Better visual feedback for interactions
- Improved loading states
- More consistent behavior across features

These changes create a more maintainable and performant foundation
for the application's UI layer while improving the overall user
experience with smoother interactions and better visual polish.

* refactor(features): modernize Skills, Prompts and Agents components

Major refactoring of feature components to improve code quality,
user experience, and maintainability.

SkillsPage Component (299 lines refactored):
- Complete rewrite of layout and state management
- Better integration with RepoManagerPanel
- Improved navigation between list and detail views
- Enhanced error handling with user-friendly messages
- Better loading states with skeleton screens
- Optimized re-renders with proper memoization
- Cleaner separation between list and form views
- Improved skill card interactions
- Better responsive design for different screen sizes

RepoManagerPanel Component (370 lines refactored):
- Streamlined repository management workflow
- Enhanced form validation with real-time feedback
- Improved repository list with better visual hierarchy
- Better handling of git operations (clone, pull, delete)
- Enhanced error recovery for network issues
- Cleaner state management reducing complexity
- Improved TypeScript type safety
- Better integration with Skills backend API
- Enhanced loading indicators for async operations

PromptPanel Component (249 lines refactored):
- Modernized layout with FullScreenPanel integration
- Better separation between list and edit modes
- Improved prompt card design with better readability
- Enhanced search and filter functionality
- Cleaner state management for editing workflow
- Better integration with PromptFormPanel
- Improved delete confirmation with safety checks
- Enhanced keyboard navigation support

PromptFormPanel Component (238 lines refactored):
- Streamlined form layout and validation
- Better markdown editor integration
- Real-time preview with syntax highlighting
- Improved validation error display
- Enhanced save/cancel workflow
- Better handling of large prompt content
- Cleaner form state management
- Improved accessibility features

AgentsPanel Component (33 lines modified):
- Minor layout adjustments for consistency
- Better integration with FullScreenPanel
- Improved placeholder states
- Enhanced error boundaries

Type Definitions (types.ts):
- Added 10 new type definitions
- Better type safety for Skills/Prompts/Agents
- Enhanced interfaces for repository management
- Improved typing for form validations

Architecture Improvements:
- Reduced component coupling
- Better prop interfaces with explicit types
- Improved error boundaries
- Enhanced code reusability
- Better testing surface

User Experience Enhancements:
- Smoother transitions between views
- Better visual feedback for actions
- Improved error messages
- Enhanced loading states
- More intuitive navigation flows
- Better responsive layouts

Code Quality:
- Net reduction of 29 lines while adding features
- Improved code organization
- Better naming conventions
- Enhanced documentation
- Cleaner control flow

These changes significantly improve the maintainability and user
experience of core feature components while establishing consistent
patterns for future development.

* style(ui): refine component layouts and improve visual consistency

Comprehensive UI polish across multiple components to enhance visual
design, improve user experience, and maintain consistency.

UsageScriptModal Component (1302 lines refactored):
- Complete layout overhaul for better usability
- Improved script editor with syntax highlighting
- Better template selection interface
- Enhanced test/preview panels with clearer separation
- Improved error feedback and validation messages
- Better modal sizing and responsiveness
- Cleaner tab navigation between sections
- Enhanced code formatting and readability
- Improved loading states for async operations
- Better integration with parent components

MCP Components:
- McpFormModal (42 lines):
  * Streamlined form layout
  * Better server type selection (stdio/http)
  * Improved field grouping and labels
  * Enhanced validation feedback
- UnifiedMcpPanel (14 lines):
  * Minor layout adjustments
  * Better list item spacing
  * Improved server status indicators
  * Enhanced action button placement

Provider Components:
- ProviderCard (11 lines):
  * Refined card layout and spacing
  * Better visual hierarchy
  * Improved badge placement
  * Enhanced hover effects
- ProviderList (5 lines):
  * Minor grid layout adjustments
  * Better drag-and-drop visual feedback
- GeminiConfigSections (4 lines):
  * Field label alignment
  * Improved spacing consistency

Editor & Footer Components:
- JsonEditor (13 lines):
  * Better editor height management
  * Improved error display
  * Enhanced syntax highlighting
- UsageFooter (10 lines):
  * Refined footer layout
  * Better quota display
  * Improved refresh button placement

Settings & Environment:
- ImportExportSection (24 lines):
  * Better button layout
  * Improved action grouping
  * Enhanced visual feedback
- EnvWarningBanner (4 lines):
  * Refined alert styling
  * Better dismiss button placement

Global Styles (index.css):
- Added 11 lines of utility classes
- Improved transition timing
- Better focus indicators
- Enhanced scrollbar styling
- Refined spacing utilities

Design Improvements:
- Consistent spacing using design tokens
- Unified color palette application
- Better typography hierarchy
- Improved shadow system for depth
- Enhanced interactive states (hover, active, focus)
- Better border radius consistency
- Refined animation timings

Accessibility:
- Improved focus indicators
- Better keyboard navigation
- Enhanced screen reader support
- Improved color contrast ratios

Code Quality:
- Net increase of 68 lines due to UsageScriptModal improvements
- Better component organization
- Cleaner style application
- Reduced style duplication

These visual refinements create a more polished and professional
interface while maintaining excellent usability and accessibility
standards across all components.

* chore(i18n): add auto-launch translation keys

Add translation keys for new auto-launch feature to support
multi-language interface.

Translation Keys Added:
- autoLaunch: Label for auto-launch toggle
- autoLaunchDescription: Explanation of auto-launch functionality
- autoLaunchEnabled: Status message when enabled

Languages Updated:
- Chinese (zh.json): 简体中文翻译
- English (en.json): English translations

The translations maintain consistency with existing terminology
and provide clear, user-friendly descriptions of the auto-launch
feature across both supported languages.

* test: update test suites to match component refactoring

Comprehensive test updates to align with recent component refactoring
and new auto-launch functionality.

Component Tests:
- AddProviderDialog.test.tsx (10 lines):
  * Updated test cases for new dialog behavior
  * Enhanced mock data for preset selection
  * Improved assertions for validation

- ImportExportSection.test.tsx (16 lines):
  * Updated for new settings page integration
  * Enhanced test coverage for error scenarios
  * Better mock state management

- McpFormModal.test.tsx (60 lines):
  * Extensive updates for form refactoring
  * New test cases for multi-app selection
  * Enhanced validation testing
  * Better coverage of stdio/http server types

- ProviderList.test.tsx (11 lines):
  * Updated for new card layout
  * Enhanced drag-and-drop testing

- SettingsDialog.test.tsx (96 lines):
  * Major updates for SettingsPage migration
  * New test cases for auto-launch functionality
  * Enhanced integration test coverage
  * Better async operation testing

Hook Tests:
- useDirectorySettings.test.tsx (32 lines):
  * Updated for refactored hook logic
  * Enhanced test coverage for edge cases

- useDragSort.test.tsx (36 lines):
  * Simplified test cases
  * Better mock implementation
  * Improved assertions

- useImportExport tests (16 lines total):
  * Updated for new error handling
  * Enhanced test coverage

- useMcpValidation.test.tsx (23 lines):
  * Updated validation test cases
  * Better coverage of error scenarios

- useProviderActions.test.tsx (48 lines):
  * Extensive updates for hook refactoring
  * New test cases for provider operations
  * Enhanced mock data

- useSettings.test.tsx (12 lines):
  * New test cases for auto-launch
  * Enhanced settings state testing
  * Better async operation coverage

Integration Tests:
- App.test.tsx (41 lines):
  * Updated for new routing logic
  * Enhanced navigation testing
  * Better component integration coverage

- SettingsDialog.test.tsx (88 lines):
  * Complete rewrite for SettingsPage
  * New integration test scenarios
  * Enhanced user workflow testing

Mock Infrastructure:
- handlers.ts (117 lines):
  * Major updates for MSW handlers
  * New handlers for auto-launch commands
  * Enhanced error simulation
  * Better request/response mocking

- state.ts (37 lines):
  * Updated mock state structure
  * New state for auto-launch
  * Enhanced state reset functionality

- tauriMocks.ts (10 lines):
  * Updated mock implementations
  * Better type safety

- server.ts & testQueryClient.ts:
  * Minor cleanup (2 lines removed)

Test Infrastructure Improvements:
- Better test isolation
- Enhanced mock data consistency
- Improved async operation testing
- Better error scenario coverage
- Enhanced integration test patterns

Coverage Improvements:
- Net increase of 195 lines of test code
- Better coverage of edge cases
- Enhanced error path testing
- Improved integration test scenarios
- Better mock infrastructure

All tests now pass with the refactored components while maintaining
comprehensive coverage of functionality and edge cases.

* style(ui): improve window dragging and provider card styles

* fix(skills): resolve third-party skills installation failure

- Add skills_path field to Skill struct
- Use skills_path to construct correct source path during installation
- Fix installation for repos with custom skill subdirectories

* feat(icon): add icon type system and intelligent inference logic

Introduce a new icon system for provider customization:

- Add IconMetadata and IconPreset interfaces in src/types/icon.ts
  - Define structure for icon name, display name, category, keywords
  - Support default color configuration per icon

- Implement smart icon inference in src/config/iconInference.ts
  - Create iconMappings for 25+ AI providers and cloud platforms
  - Include Claude, DeepSeek, Qwen, Kimi, Google, AWS, Azure, etc.
  - inferIconForPreset(): match provider name to icon config
  - addIconsToPresets(): batch apply icons to preset arrays
  - Support fuzzy matching for flexible name recognition

This foundation enables automatic icon assignment when users add
providers, improving visual identification in the provider list.

* feat(ui): add icon picker, color picker and provider icon components

Implement comprehensive icon selection system for provider customization:

## New Components

### ProviderIcon (src/components/ProviderIcon.tsx)
- Render SVG icons by name with automatic fallback
- Display provider initials when icon not found
- Support custom sizing via size prop
- Use dangerouslySetInnerHTML for inline SVG rendering

### IconPicker (src/components/IconPicker.tsx)
- Grid-based icon selection with visual preview
- Real-time search filtering by name and keywords
- Integration with icon metadata for display names
- Responsive grid layout (6-10 columns based on screen)

### ColorPicker (src/components/ColorPicker.tsx)
- 12 preset colors for quick selection
- Native color input for custom color picking
- Hex input field for precise color entry
- Visual feedback for selected color state

## Icon Assets (src/icons/extracted/)
- 38 high-quality SVG icons for AI providers and platforms
- Includes: OpenAI, Claude, DeepSeek, Qwen, Kimi, Gemini, etc.
- Cloud platforms: AWS, Azure, Google Cloud, Cloudflare
- Auto-generated index.ts with getIcon/hasIcon helpers
- Metadata system with searchable keywords per icon

## Build Scripts
- scripts/extract-icons.js: Extract icons from simple-icons
- scripts/generate-icon-index.js: Generate TypeScript index file

* feat(provider): integrate icon system into provider UI components

Add icon customization support to provider management interface:

## Type System Updates

### Provider Interface (src/types.ts)
- Add optional `icon` field for icon name (e.g., "openai", "anthropic")
- Add optional `iconColor` field for hex color (e.g., "#00A67E")

### Form Schema (src/lib/schemas/provider.ts)
- Extend providerSchema with icon and iconColor optional fields
- Maintain backward compatibility with existing providers

## UI Components

### ProviderCard (src/components/providers/ProviderCard.tsx)
- Display ProviderIcon alongside provider name
- Add icon container with hover animation effect
- Adjust layout spacing for icon placement
- Update translate offsets for action buttons

### BasicFormFields (src/components/providers/forms/BasicFormFields.tsx)
- Add icon preview section showing current selection
- Implement fullscreen icon picker dialog
- Auto-apply default color from icon metadata on selection
- Display provider name and icon status in preview

### AddProviderDialog & EditProviderDialog
- Pass icon fields through form submission
- Preserve icon data during provider updates

This enables users to visually distinguish providers in the list
with custom icons, improving UX for multi-provider setups.

* feat(backend): add icon fields to Provider model and default mappings

Extend Rust backend to support provider icon customization:

## Provider Model (src-tauri/src/provider.rs)
- Add `icon: Option<String>` field for icon name
- Add `icon_color: Option<String>` field for hex color
- Use serde rename `iconColor` for frontend compatibility
- Apply skip_serializing_if for clean JSON output
- Update Provider::new() to initialize icon fields as None

## Provider Defaults (src-tauri/src/provider_defaults.rs) [NEW]
- Define ProviderIcon struct with name and color fields
- Create DEFAULT_PROVIDER_ICONS static HashMap with 23 providers:
  - AI providers: OpenAI, Anthropic, Claude, Google, Gemini,
    DeepSeek, Kimi, Moonshot, Zhipu, MiniMax, Baidu, Alibaba,
    Tencent, Meta, Microsoft, Cohere, Perplexity, Mistral, HuggingFace
  - Cloud platforms: AWS, Azure, Huawei, Cloudflare
- Implement infer_provider_icon() with exact and fuzzy matching
- Add unit tests for matching logic (exact, fuzzy, case-insensitive)

## Deep Link Support (src-tauri/src/deeplink.rs)
- Initialize icon fields when creating Provider from deep link import

## Module Registration (src-tauri/src/lib.rs)
- Register provider_defaults module

## Dependencies (Cargo.toml)
- Add once_cell for lazy static initialization

This backend support enables icon persistence and future features
like auto-icon inference during provider creation.

* chore(i18n): add translations for icon picker and provider icon

Add Chinese and English translations for icon customization feature:

## Icon Picker (iconPicker)
- search: "Search Icons" / "搜索图标"
- searchPlaceholder: "Enter icon name..." / "输入图标名称..."
- noResults: "No matching icons found" / "未找到匹配的图标"
- category.aiProvider: "AI Providers" / "AI 服务商"
- category.cloud: "Cloud Platforms" / "云平台"
- category.tool: "Dev Tools" / "开发工具"
- category.other: "Other" / "其他"

## Provider Icon (providerIcon)
- label: "Icon" / "图标"
- colorLabel: "Icon Color" / "图标颜色"
- selectIcon: "Select Icon" / "选择图标"
- preview: "Preview" / "预览"

These translations support the new icon picker UI components
and provider form icon selection interface.

* style(ui): refine header layout and AppSwitcher color scheme

Improve application header and component styling:

## App.tsx Header Layout
- Wrap title and settings button in flex container with gap
- Add vertical divider between title and settings icon
- Apply responsive padding (pl-1 sm:pl-2)
- Reformat JSX for better readability (prettier)
- Fix string template formatting in className

## AppSwitcher Color Update
- Change Claude tab gradient from orange/amber to teal/emerald/green
- Update shadow color to match new teal theme
- Change hover color from orange-500 to teal-500
- Align visual style with emerald/teal brand colors

## Dialog Component Cleanup
- Remove default close button (X icon) from DialogContent
- Allow parent components to control close button placement
- Remove unused lucide-react X import

## index.css Header Border
- Add top border (2px solid) to glass-header
- Apply to both light and dark mode variants
- Improve visual separation of header area

These changes enhance visual consistency and modernize the UI
appearance with a cohesive teal color scheme.

* chore(deps): add icon library and update preset configurations

Add dependencies and utility scripts for icon system:

## Dependencies (package.json)
- Add @lobehub/icons-static-svg@1.73.0
  - High-quality SVG icon library for AI providers
  - Source for extracted icons in src/icons/extracted/
- Update pnpm-lock.yaml accordingly

## Provider Preset Updates (src/config/claudeProviderPresets.ts)
- Add optional `icon` and `iconColor` fields to ProviderPreset interface
- Apply to Anthropic Official preset as example:
  - icon: "anthropic"
  - iconColor: "#D4915D"
- Future presets can include default icon configurations

## Utility Script (scripts/filter-icons.js) [NEW]
- Helper script for filtering and managing icon assets
- Supports icon discovery and validation workflow
- Complements extract-icons.js and generate-icon-index.js

This completes the icon system infrastructure, providing all
necessary tools and dependencies for icon customization.

* refactor(ui): simplify AppSwitcher styles and migrate to local SVG icons

- Replace complex gradient animations with clean, minimal tab design
- Migrate from @lobehub/icons CDN to local SVG assets for better reliability
- Fix clippy warning in error.rs (use inline format args)
- Improve code formatting in skill service and commands
- Reduce CSS complexity in AppSwitcher component (removed blur effects and gradients)
- Update BrandIcons to use imported local SVG files instead of dynamic image loading

This improves performance, reduces external dependencies, and provides a cleaner UI experience.

* style(ui): hide scrollbars across all browsers and optimize form layout

- Hide scrollbars globally with cross-browser support:
  * WebKit browsers (Chrome, Safari, Edge): ::-webkit-scrollbar { display: none }
  * Firefox: scrollbar-width: none
  * IE 10+: -ms-overflow-style: none
- Remove custom scrollbar styling (width, colors, hover states)
- Reorganize BasicFormFields layout:
  * Move icon picker to top center as a clickable preview (80x80)
  * Change name and notes fields to horizontal grid layout (md:grid-cols-2)
  * Remove icon preview section from bottom
  * Improve visual hierarchy and space efficiency

This provides a cleaner, more modern UI with hidden scrollbars while maintaining full scroll functionality.

* refactor(layout): standardize max-width to 60rem and optimize padding structure

- Unify container max-width across components:
  * Replace max-w-4xl with max-w-[60rem] in App.tsx provider list
  * Replace max-w-5xl with max-w-[60rem] in PromptPanel
  * Move padding from header element to inner container for consistent spacing
- Optimize padding hierarchy:
  * Remove px-6 from header element, add to inner header container
  * Remove px-6 from main element, keep it only in provider list container
  * Consolidate PromptPanel padding: move px-6 from nested divs to outer container
  * Remove redundant pl-1 sm:pl-2 from logo/title area
- Benefits:
  * Consistent 60rem max-width provides better readability on wide screens
  * Simplified padding structure reduces CSS complexity
  * Cleaner responsive behavior with unified spacing rules

This creates a more maintainable and visually consistent layout system.

* refactor(ui): unify layout system with 60rem max-width and consistent spacing

- Standardize container max-width across all panels:
  * Replace max-w-4xl and max-w-5xl with unified max-w-[60rem]
  * Apply to SettingsPage, UnifiedMcpPanel, SkillsPage, and FullScreenPanel
  * Ensures consistent reading width and visual balance on wide screens

- Optimize padding hierarchy and structure:
  * Move px-6 from parent elements to content containers
  * FullScreenPanel: Add max-w-[60rem] wrapper to header, content, and footer
  * Add border separators (border-t/border-b) to header and footer sections
  * Consolidate nested padding in MCP, Skills, and Prompts panels
  * Remove redundant padding layers for cleaner CSS

- Simplify component styling:
  * MCP list items: Replace card-based layout with modern group-based design
  * Remove unnecessary wrapper divs and flatten DOM structure
  * Update card hover effects with smooth transitions
  * Simplify icon selection dialog (remove description text in BasicFormFields)

- Benefits:
  * Consistent 60rem max-width provides optimal readability
  * Unified spacing rules reduce maintenance complexity
  * Cleaner component hierarchy improves performance
  * Better responsive behavior across different screen sizes
  * More cohesive visual design language throughout the app

This creates a maintainable, scalable design system foundation.

* feat(deeplink): add Claude model fields support and enhance import dialog

- Add Claude-specific model field support in deeplink import:
  * Support model (ANTHROPIC_MODEL) - general default model
  * Support haikuModel (ANTHROPIC_DEFAULT_HAIKU_MODEL)
  * Support sonnetModel (ANTHROPIC_DEFAULT_SONNET_MODEL)
  * Support opusModel (ANTHROPIC_DEFAULT_OPUS_MODEL)
  * Backend: Update DeepLinkImportRequest struct to include optional model fields
  * Frontend: Add TypeScript type definitions for new model parameters

- Enhance deeplink demo page (deplink.html):
  * Add 5 new Claude configuration examples showcasing different model setups
  * Add parameter documentation with required/optional tags
  * Include basic config (no models), single model, complete 4-model, partial models, and third-party provider examples
  * Improve visual design with param-list component and color-coded badges
  * Add detailed descriptions for each configuration scenario

- Redesign DeepLinkImportDialog layout:
  * Switch from 3-column to compact 2-column grid layout
  * Reduce dialog width from 500px to 650px for better content display
  * Add dedicated section for Claude model configurations with blue highlight box
  * Use uppercase labels and smaller text for more information density
  * Add truncation and tooltips for long URLs
  * Improve visual hierarchy with spacing and grouping
  * Increase z-index to 9999 to ensure dialog appears on top

- Minor UI refinements:
  * Update App.tsx layout adjustments
  * Optimize McpFormModal styling
  * Refine ProviderCard and BasicFormFields components

This enables users to import Claude providers with precise model configurations via deeplink.

* feat(deeplink): add config file support for deeplink import

Support importing provider configuration from embedded or remote config files.
- Add base64 dependency for config content encoding
- Support config, configFormat, and configUrl parameters
- Make homepage/endpoint/apiKey optional when config is provided
- Add config parsing and merging logic

* feat(deeplink): enhance dialog with config file preview

Add config file parsing and preview in deep link import dialog.
- Support Base64 encoded config display
- Add config file source indicator (embedded/remote)
- Parse and display config fields by app type (Claude/Codex/Gemini)
- Mask sensitive values in config preview
- Improve dialog layout and content organization

* refactor(ui): unify dialog styles and improve layout consistency

Standardize dialog and panel components across the application.
- Update dialog background to use semantic color tokens
- Adjust FullScreenPanel max-width to 56rem for better alignment
- Add drag region and prevent body scroll in full-screen panels
- Optimize button sizes and spacing in panel headers
- Apply consistent styling to all dialog-based components

* i18n: add deeplink config preview translations

Add missing translation keys for config file preview feature.
- Add configSource, configEmbedded, configRemote labels
- Add configDetails and configUrl display strings
- Support both Chinese and English versions

* feat(deeplink): enhance test page with v3.8 config file examples

Improve deeplink test page with comprehensive config file import examples.
- Add version badge for v3.8 features
- Add copy-to-clipboard functionality for all deep links
- Add Claude config file import examples (embedded/remote)
- Add Codex config file import examples (auth.json + config.toml)
- Add Gemini config file import examples (.env format)
- Add config generator tool for easy testing
- Update UI with better styling and layout

* feat(settings): add autoSaveSettings for lightweight auto-save

Add optimized auto-save function for General tab settings.
- Add autoSaveSettings method for non-destructive auto-save
- Only trigger system APIs when values actually change
- Avoid unnecessary auto-launch and plugin config updates
- Update tests to cover new functionality

* refactor(settings): simplify settings page layout and auto-save

Reorganize settings page structure and integrate autoSaveSettings.
- Move save button inline within Advanced tab content
- Remove sticky footer for cleaner layout
- Use autoSaveSettings for General tab settings
- Simplify dialog close behavior
- Refactor ImportExportSection layout

* style(providers): optimize card layout and action button sizes

Improve provider card visual density and action buttons.
- Reduce icon button sizes for compact layout
- Adjust drag handle and icon sizes
- Tighten spacing between action buttons
- Update hover translate values for better alignment

* refactor(mcp): improve form modal layout with adaptive height editor

Restructure MCP form modal for better space utilization.
- Split form into upper form fields and lower JSON editor sections
- Add full-height mode support for JsonEditor component
- Use flex layout for editor to fill available space
- Update PromptFormPanel to use full-height editor
- Fix locale text formatting

* style: unify list item styles with semantic colors

Apply consistent styling to list items across components.
- Replace hardcoded colors with semantic tokens in MCP and Prompt list items
- Add glass effect container to EndpointSpeedTest panel
- Format code for better readability

* style: format template literals for better readability

Improve code formatting for conditional className expressions.
- Break long template literals across multiple lines
- Maintain consistent formatting in MCP form and endpoint test components

* feat(deeplink): add config merge command for preview

Expose config merging functionality to frontend for preview.
- Add merge_deeplink_config Tauri command
- Make parse_and_merge_config public for reuse
- Enable frontend to display complete config before import

* feat(deeplink): merge and display config in import dialog

Enhance import dialog to fetch and display complete config.
- Call mergeDeeplinkConfig API when config is present
- Add UTF-8 base64 decoding support for config content
- Add scrollable content area with custom scrollbar styling
- Show complete configuration before user confirms import

* i18n: add config merge error message

Add translation for config file merge error handling.

* style(deeplink): format test page HTML for better readability

Improve HTML formatting in deeplink test page.
- Format multiline attributes for better readability
- Add consistent indentation to nested elements
- Break long lines in buttons and links

* refactor(usage): improve footer layout with two-row design

Reorganize usage footer for better readability and space efficiency.
- Split into two rows: update time + refresh button (row 1), usage stats (row 2)
- Move refresh button to top row next to update time
- Remove card background for cleaner look
- Add fallback text when never updated
- Improve spacing and alignment
- Format template literals for consistency

* feat(database): add SQLite database infrastructure

- Add rusqlite dependency (v0.32.1) and r2d2 connection pooling
- Implement Database module with CRUD operations for providers, MCP servers, prompts, and skills
- Add schema initialization with proper indexes
- Include data migration utilities from JSON config to SQLite
- Support timestamp tracking (created_at, updated_at)

* refactor(core): integrate SQLite database into application core

- Initialize database on app startup with migration from JSON config
- Update AppState to include Database instance alongside MultiAppConfig
- Simplify store module by removing unused session management code
- Add database initialization to app setup flow
- Support both database and legacy config during transition

* refactor(services): migrate service layer to use SQLite database

- Refactor ProviderService to use database queries instead of in-memory config
- Update McpService to fetch and store MCP servers in database
- Migrate PromptService to database-backed storage
- Simplify ConfigService by removing complex transaction logic
- Remove 648 lines of redundant code through database abstraction

* refactor(commands): update command layer to use database API

- Update config commands to query database for providers and settings
- Modify provider commands to pass database handle to services
- Update MCP commands to use database-backed operations
- Refactor prompt and skill commands to leverage database storage
- Simplify import/export commands with database integration

* refactor(backend): update supporting modules for database compatibility

- Add DatabaseError variant to AppError enum
- Update provider module to support database-backed operations
- Modify codex_config to work with new database structure
- Ensure error handling covers database operations

* refactor(frontend): update UI components for database migration

- Update UsageFooter component to handle new data structure
- Modify SkillsPage to work with database-backed skills management
- Ensure frontend compatibility with refactored backend

* feat(skills): add search functionality to Skills page

- Add search input with Search icon in SkillsPage component
- Implement useMemo-based filtering by skill name, description, and directory
- Display search results count when filtering is active
- Show "no results" message when no skills match the search query
- Add i18n translations for search UI (zh/en)
- Maintain responsive layout and consistent styling with existing UI

* refactor(backend): replace unsafe unwrap calls with proper error handling

- Add to_json_string helper for safe JSON serialization
- Add lock_conn macro for safe Mutex locking
- Replace 41 unwrap() calls with proper error handling:
  - database.rs: JSON serialization and Mutex operations (31 fixes)
  - lib.rs: macOS NSWindow and tray icon handling (3 fixes)
  - services/provider.rs: Claude model normalization (1 fix)
  - services/prompt.rs: timestamp generation (3 fixes)
  - services/skill.rs: directory name extraction (2 fixes)
  - mcp.rs: HashMap initialization and type conversions (5 fixes)
  - app_config.rs: timestamp fallback (1 fix)

This improves application stability and prevents potential panics.

* feat(init): implement automatic data import on first launch

Add comprehensive first-launch data import system:

Database layer:
- Add is_empty_for_first_import() to detect empty database
- Add init_default_skill_repos() to initialize 3 default skill repositories

Services layer:
- Implement McpService::import_from_claude/codex/gemini()
  to import MCP servers from existing config files
- Implement PromptService::import_from_file_on_first_launch()
  to import prompt files (CLAUDE.md, AGENTS.md, GEMINI.md)

Startup flow (lib.rs):
- Check if database is empty on startup
- Import existing configurations if detected:
  1. Initialize default skill repositories
  2. Import provider configurations from live settings
  3. Import MCP servers from config files
  4. Import prompt files
- All imports are fault-tolerant and logged

This ensures seamless migration from file-based configs to database.

* fix(skills): auto-sync locally installed skills to database

Add automatic synchronization in get_skills command:
- Detect locally installed skills in ~/.claude/skills/
- Auto-add to database if not already tracked
- Ensures existing skills are recognized on first launch

This fixes the issue where user's existing skills were not
imported into the database on initial application run.

* docs(frontend): add code comments and improve formatting

- Add explanatory comment in EditProviderDialog about config assembly
- Improve import formatting in SkillsPage for better readability

* feat(deeplink): display all four Claude model fields in import dialog

- Show haiku/sonnet/opus/multiModel fields conditionally for Claude
- Maintain single model field display for Codex and Gemini
- Add i18n translations for new model field labels (zh/en)

* feat(backend): add database SQL export/import with backup

- Enable rusqlite backup feature for SQL dump support
- Implement export_sql to generate SQLite-compatible SQL dumps
- Implement import_sql with automatic backup before import
- Add snapshot_to_memory to avoid long-held database locks
- Add backup rotation to retain latest 10 backups
- Support atomic import with rollback on failure

* refactor(backend): migrate import/export to use SQL backup

- Reimplement export_config_to_file to use database.export_sql
- Reimplement import_config_from_file to use database.import_sql
- Add sync_current_from_db to sync live configs after import
- Add settings database binding on app initialization
- Remove deprecated JSON-based config import logic

* refactor(backend): migrate settings storage to database

- Add bind_db function to initialize database-backed settings
- Implement load_initial_settings with database fallback
- Replace direct file save with settings store management
- Add SETTINGS_DB static for database binding
- Maintain backward compatibility with file-based settings
- Keep settings.json for legacy migration support

* feat(frontend): update import/export UI for SQL backup

- Change default export filename from .json to .sql
- Update file format to timestamp format (YYYYMMDD_HHMMSS)
- Update error messages to reference SQL files
- Align with backend SQL export/import implementation

* feat(i18n): update translations for SQL backup feature

- Update Chinese translations for SQL import/export
- Update English translations for SQL import/export
- Change terminology from 'config file' to 'SQL backup'
- Update error messages and UI hints

* fix(backend): remove unnecessary dereference in backup operation

- Simplify Backup::new call by removing redundant dereference
- MutexGuard already implements DerefMut

* feat(icons): add PackyCode provider icon support

Add PackyCode as a supported AI provider icon with proper metadata
and filtering configuration.

Changes:
- Add 'packycode' to icon filter whitelist in filter-icons.js
- Register PackyCode metadata with display name, category, and keywords
- Import PackyCode SVG icon file
- Export icon through index.ts for use in provider configurations

The PackyCode icon uses currentColor to adapt to theme styling.

* feat(utils): add base64 encoding utility functions

Add reusable base64 encoding/decoding utility functions for handling
binary data and string conversions in deeplink imports.

Features:
- encodeBase64: Encode string to base64
- decodeBase64: Decode base64 to string
- Uses browser-native btoa/atob with proper UTF-8 handling

This utility will be used for encoding prompt content and configuration
files in deeplink URLs.

* feat(backend): add in-memory database mode for testing

Add support for creating in-memory SQLite database instances to
improve test isolation and performance.

Changes:
- Add Database::memory() constructor for in-memory database
- Enable foreign key constraints for data integrity
- Export Database type from lib.rs for test usage
- Initialize tables automatically on memory database creation

This enables unit tests to run without filesystem dependencies and
provides faster test execution with proper cleanup.

* refactor(deeplink): extend support for multi-resource imports

Extend the deeplink import system to support importing multiple
resource types beyond providers: prompts, MCP servers, and skills.

Breaking changes:
- DeepLinkImportRequest: Convert required fields to Optional to
  support different resource types (app, name, homepage, endpoint,
  apiKey are now Option<String>)
- Add resource-specific fields for prompt, mcp, and skill imports

New features:
- parse_prompt_deeplink: Parse prompt import URLs with base64 content
- parse_mcp_deeplink: Parse MCP server import URLs with config
- parse_skill_deeplink: Parse GitHub skill repository URLs
- import_prompt_from_deeplink: Import prompts to database
- import_mcp_from_deeplink: Batch import MCP servers with multi-app support
- import_skill_from_deeplink: Clone and install skill repositories

Data model additions:
- Prompt fields: content (base64), description, enabled
- MCP fields: apps (comma-separated), config, config_format
- Skill fields: repo, directory, branch, skills_path
- Common fields: icon (provider icon name)

McpImportResult type:
- imported_count: Number of successfully imported servers
- imported_ids: List of imported server IDs
- failed: List of failed imports with error messages

URL format examples:
- Prompt: ccswitch://v1/import?resource=prompt&app=claude&name=...&content=...
- MCP: ccswitch://v1/import?resource=mcp&apps=claude,codex&config=...
- Skill: ccswitch://v1/import?resource=skill&repo=owner/name&directory=...

This refactor enables one-click sharing of prompts, MCP configurations,
and skill repositories via deeplink URLs.

* feat(backend): add unified deeplink import command

Add a new unified command handler for importing all resource types
via deeplinks, replacing the legacy provider-only import command.

Changes:
- Add import_from_deeplink_unified command supporting all resource types
- Keep import_from_deeplink for backward compatibility (now marked legacy)
- Route imports based on request.resource field (provider/prompt/mcp/skill)
- Return typed ImportResult with resource-specific data

Return types:
- Provider: { type: "provider", id: string }
- Prompt: { type: "prompt", id: string }
- MCP: { type: "mcp", importedCount, importedIds, failed }
- Skill: { type: "skill", key: string }

The unified handler simplifies frontend logic by providing consistent
return types and error handling across all resource types.

* feat(frontend): extend deeplink API for multi-resource support

Update the frontend deeplink API to support importing multiple
resource types with proper TypeScript typing.

Changes:
- Add ResourceType union type: "provider" | "prompt" | "mcp" | "skill"
- Convert DeepLinkImportRequest fields to optional (matching backend)
- Add resource-specific field types (prompt, mcp, skill)
- Add ImportResult discriminated union for type-safe results
- Add McpImportResult interface for batch import results
- Update importFromDeeplink to use unified command

Type safety improvements:
- ImportResult discriminated union ensures proper type narrowing
- Each result type has its own specific return data structure
- Frontend can pattern match on result.type for correct handling

Breaking change:
- importFromDeeplink now returns ImportResult instead of string
- Callers must handle all resource types appropriately

* feat(frontend): add resource-specific confirmation dialog components

Add specialized confirmation UI components for each deeplink import
resource type (Prompt, MCP, Skill).

Components:
- PromptConfirmation: Display prompt name, app, description, and
  content preview with markdown rendering
- McpConfirmation: Show MCP server list with target apps, supports
  batch import display
- SkillConfirmation: Display GitHub repository info with branch and
  directory details

Features:
- Consistent card-based layout with proper spacing
- Sensitive data masking (API keys shown as dots)
- Icon support for providers (ProviderIcon component)
- Badge components for visual status indicators
- Responsive design with proper text overflow handling

Each component focuses on displaying the most relevant information
for users to make informed import decisions.

* feat(frontend): update DeepLinkImportDialog for multi-resource imports

Refactor the main deeplink import dialog to handle all resource types
with proper confirmation UI and post-import actions.

Key changes:
- Add resource-specific confirmation components (Prompt/MCP/Skill)
- Implement typed result handling with discriminated unions
- Add MCP result type guard for backward compatibility
- Implement resource-specific cache invalidation strategies
- Add custom event dispatching for non-React-Query resources

Import flow improvements:
- Provider: Invalidate provider queries, show success toast
- Prompt: Dispatch "prompt-imported" event, trigger manual refresh
- MCP: Aggressive cache invalidation with refetchQueries, handle
  partial success (show warning if some imports failed)
- Skill: Force refetch skills query with refetchType: "all"

Error handling:
- Graceful fallback for legacy backend responses (no type field)
- MCP-specific result detection via type guard
- Detailed error messages in toasts

UI improvements:
- Dynamic dialog title based on resource type
- Resource-specific confirmation content rendering
- Better visual feedback during import process

* feat(frontend): add deeplink import event listeners and UI improvements

Add event-driven refresh logic for deeplink imports and enhance
Skills page filtering capabilities.

PromptPanel changes:
- Add "prompt-imported" custom event listener
- Auto-reload prompts when deeplink import completes
- Filter events by app ID to avoid unnecessary refreshes
- Clean up event listener on component unmount

SkillsPage improvements:
- Add installation status filter (all/installed/uninstalled)
- Implement Select component for filter dropdown
- Combine status filter with existing search functionality
- Update filtered skills memo to include both filters
- Improve responsive layout for search and filter controls

Event flow:
1. DeepLinkImportDialog dispatches "prompt-imported" event
2. PromptPanel listens for event matching its app
3. Panel triggers reload to show newly imported prompt
4. Similar pattern can be used for other non-React-Query resources

These improvements enable seamless UI updates after deeplink imports
without requiring manual page refresh.

* test(deeplink): migrate tests to use in-memory database

Update deeplink import tests to use the new in-memory database
instead of filesystem-based configuration.

Changes:
- Replace MultiAppConfig with Database-based AppState
- Use Database::memory() for isolated test instances
- Update provider verification to query database directly
- Add icon field verification in test assertions
- Remove filesystem config.json validation (now DB-backed)

Test improvements:
- Faster execution (no disk I/O)
- Better isolation (each test gets fresh DB instance)
- No cleanup required (memory DB auto-discarded)
- Consistent with v3.8+ storage architecture

Updated tests:
- deeplink_import_claude_provider_persists_to_db
- deeplink_import_codex_provider_builds_auth_and_config

Both tests verify that deeplink imports correctly persist provider
data to the database with all expected fields.

* feat(i18n): add translations for deeplink and skills features

Add internationalization support for new deeplink import features
and skills page filtering.

Deeplink translations:
- Add "icon" field label for provider icon selection
- Both Chinese ("图标") and English ("Icon") translations

Skills page translations:
- Add filter placeholder and options
- Filter states: "all", "installed", "uninstalled"
- Chinese: "全部", "已安装", "未安装"
- English: "All", "Installed", "Not installed"

These translations ensure consistent multilingual support for the
new multi-resource deeplink import system and improved skills
management UI.

* chore: add deeplink testing HTML page

Add a local HTML page for testing deeplink protocol functionality
during development.

This page allows developers to:
- Test different deeplink URL formats (provider/prompt/mcp/skill)
- Verify URL parsing and parameter encoding
- Quickly validate deeplink imports without external tools
- Debug protocol registration on different platforms

Not intended for production use, only for development testing.

* refactor(ui): improve icon rendering consistency and spacing

Standardize icon sizes and improve rendering consistency across the
application interface.

Changes:
- ProviderIcon: Add fontSize sync with size prop for embedded SVG
  scaling, improve fallback text sizing calculation
- AppSwitcher: Replace brand icon components with unified ProviderIcon,
  standardize icon size to 20px across all app tabs
- ProviderCard: Reduce icon size from 26px to 20px for better visual
  balance in card layout
- DeepLinkImportDialog: Increase confirmation icon size from 64px to
  80px for better visibility
- BasicFormFields: Reduce icon picker button spacing from space-y-6 to
  space-y-2 for improved layout density

Icon rendering improvements:
- Embedded SVGs now properly scale with fontSize set to match size prop
- Fallback initials use responsive font sizing (50% of icon size,
  minimum 12px)
- Consistent 20px standard for navigation and card icons
- Better visual hierarchy with appropriate sizing for different contexts

This refactor improves visual consistency and makes icon rendering
more predictable across different components.
2025-11-25 09:30:55 +08:00
Jason 1af20b5f8f feat: add dry-run mode for JSON→SQLite migration
Core changes:
- Extract transaction logic into reusable migrate_from_json_tx() helper
- Add migrate_from_json_dry_run() for in-memory validation without disk writes
- Implement three-state migration mode enum (Disabled/DryRun/Enabled)
- Support CC_SWITCH_ENABLE_JSON_DB_MIGRATION=dryrun for safe testing

Code quality improvements:
- Remove redundant migration_needed variable
- Unify all log messages to English
- Use info level for user-initiated operations instead of warn
- Add explicit drop(tx) in dry-run to clarify intent

Testing:
- Add dry_run_does_not_write_to_disk test
- Add dry_run_validates_schema_compatibility test with real data
- All 6 database tests passing, zero clippy warnings

This enables users to safely validate migration compatibility before
committing to the database, catching schema mismatches early.
2025-11-24 23:35:39 +08:00
Jason ea54b7d010 fix: improve database schema handling and add migration feature gate
- Fix column name quoting in ALTER TABLE statements to prevent SQL
  syntax errors with reserved keywords
- Use case-insensitive table name matching in table_exists()
- Change type declarations from INTEGER to BOOLEAN for semantic clarity
- Add CC_SWITCH_ENABLE_JSON_DB_MIGRATION env var to gate JSON→SQLite
  migration (disabled by default until feature is stable)
- Refactor tests with shared legacy schema and column info helpers
- Add comprehensive test for column types and default values alignment
2025-11-24 22:52:58 +08:00
Jason f93b21c97e fix: improve database migration robustness and schema consistency
This commit addresses several critical issues in the database migration system:

**Schema Consistency**
- Unified DEFAULT and NOT NULL constraints between CREATE TABLE and ALTER TABLE
- Fixed inconsistencies in: providers.meta, mcp_servers.tags, skills.installed_at, skill_repos.branch
- Ensures new installations and migrations produce identical schemas

**Security & Validation**
- Added SQL identifier validation to prevent potential injection risks
- Validates table and column names contain only alphanumeric characters and underscores

**Error Handling**
- Fixed table_exists() to distinguish "table not found" from "query failed"
- Properly propagates database errors instead of silently ignoring them
- Uses pattern matching on rusqlite::Error::QueryReturnedNoRows

**Transaction Safety**
- Implemented SAVEPOINT-based transaction protection for migrations
- Ensures user_version stays consistent with actual schema on failures
- Graceful rollback on migration errors (within SQLite's ALTER TABLE limitations)

**Testing**
- Enhanced test coverage to verify NOT NULL constraints and DEFAULT values
- Validates that migration produces the expected schema structure

These improvements ensure database migrations are more reliable, secure, and maintainable.
2025-11-24 22:34:33 +08:00
Jason a7ca6fb985 feat: add schema version management for database migrations
Implement SQLite PRAGMA user_version based migration system:
- Track schema version with SCHEMA_VERSION constant
- Apply migrations automatically on init and import
- Reject databases from future versions for forward compatibility
- Add comprehensive tests for version transitions
- Prepare infrastructure for future schema evolution

This lays the foundation for safe incremental database upgrades.
2025-11-24 17:06:26 +08:00
Jason 67aa275599 test: migrate tests to SQLite database architecture
This commit refactors all tests to work with the new database-based
architecture, replacing the previous JSON config approach.

Key changes:
- Add Database export to lib.rs for test access
- Create test helper functions in support.rs:
  - create_test_state(): Creates empty test state with fresh DB
  - create_test_state_with_config(): Migrates JSON config to DB
- Fix environment isolation in provider_service tests:
  - provider_service_switch_missing_provider_returns_error
  - provider_service_switch_codex_missing_auth_returns_error
- Replace ignored export tests with working alternatives:
  - export_sql_writes_to_target_path (tests Database::export_sql)
  - export_sql_returns_error_for_invalid_path (tests error handling)
- Update error type matching to align with current implementation

All tests now:
- Use isolated test environments (test_mutex + reset_test_fs)
- Access data via Database API instead of RwLock<MultiAppConfig>
- Work with SQLite persistence layer
- Pass without environment pollution or race conditions

Fixes test compilation errors after database migration.
2025-11-24 12:24:41 +08:00
Jason fe190081eb fix: resolve critical bugs in settings and import flow
Fixed two critical issues:

1. **Blocking Issue - restart_app return type error**
   - Fixed compilation error where restart_app() didn't return a value
   - Used async spawn with 100ms delay to allow response before restart
   - Prevents "unreachable code" compiler error

2. **High Priority - Import SQL doesn't refresh AppSettings cache**
   - Added reload_settings() function to refresh in-memory settings cache
   - Integrated into import flow to ensure imported settings take effect
   - Prevents imported settings being overwritten by stale memory cache
   - Affects: language, config directories, auto-launch, custom endpoints

Changes:
- src/commands/settings.rs: Async delayed restart with proper return value
- src/settings.rs: New reload_settings() to sync memory cache from DB
- src/commands/import_export.rs: Call reload_settings() after SQL import

Verified: cargo clippy --lib and pnpm typecheck both pass
2025-11-24 11:25:41 +08:00
YoVinchen d30562954a Refactor/storage (#277)
* feat(components): add reusable full-screen panel components

Add new full-screen panel components to support the UI refactoring:

- FullScreenPanel: Reusable full-screen layout component with header,
  content area, and optional footer. Provides consistent layout for
  settings, prompts, and other full-screen views.

- PromptFormPanel: Dedicated panel for creating and editing prompts
  with markdown preview support. Features real-time validation and
  integrated save/cancel actions.

- AgentsPanel: Panel component for managing agent configurations.
  Provides a consistent interface for agent CRUD operations.

- RepoManagerPanel: Full-featured repository manager panel for Skills.
  Supports repository listing, addition, deletion, and configuration
  management with integrated validation.

These components establish the foundation for the upcoming settings
page migration from dialog-based to full-screen layout.

* refactor(settings): migrate from dialog to full-screen page layout

Complete migration of settings from modal dialog to dedicated full-screen
page, improving UX and providing more space for configuration options.

Changes:
- Remove SettingsDialog component (legacy modal-based interface)
- Add SettingsPage component with full-screen layout using FullScreenPanel
- Refactor App.tsx routing to support dedicated settings page
  * Add settings route handler
  * Update navigation logic from dialog-based to page-based
  * Integrate with existing app switcher and provider management
- Update ImportExportSection to work with new page layout
  * Improve spacing and layout for better readability
  * Enhanced error handling and user feedback
  * Better integration with page-level actions
- Enhance useSettings hook to support page-based workflow
  * Add navigation state management
  * Improve settings persistence logic
  * Better error boundary handling

Benefits:
- More intuitive navigation with dedicated settings page
- Better use of screen space for complex configurations
- Improved accessibility with clearer visual hierarchy
- Consistent with modern desktop application patterns
- Easier to extend with new settings sections

This change is part of the larger UI refactoring initiative to modernize
the application interface and improve user experience.

* refactor(forms): simplify and modernize form components

Comprehensive refactoring of form components to reduce complexity,
improve maintainability, and enhance user experience.

Provider Forms:
- CodexCommonConfigModal & CodexConfigSections
  * Simplified state management with reduced boilerplate
  * Improved field validation and error handling
  * Better layout with consistent spacing
  * Enhanced model selection with visual indicators
- GeminiCommonConfigModal & GeminiConfigSections
  * Streamlined authentication flow (OAuth vs API Key)
  * Cleaner form layout with better grouping
  * Improved validation feedback
  * Better integration with parent components
- CommonConfigEditor
  * Reduced from 178 to 68 lines (-62% complexity)
  * Extracted reusable form patterns
  * Improved JSON editing with syntax validation
  * Better error messages and recovery options
- EndpointSpeedTest
  * Complete rewrite for better UX
  * Real-time testing progress indicators
  * Enhanced error handling with retry logic
  * Visual feedback for test results (color-coded latency)

MCP & Prompts:
- McpFormModal
  * Simplified from 581 to ~360 lines
  * Better stdio/http server type handling
  * Improved form validation
  * Enhanced multi-app selection (Claude/Codex/Gemini)
- PromptPanel
  * Cleaner integration with PromptFormPanel
  * Improved list/grid view switching
  * Better state management for editing workflows
  * Enhanced delete confirmation with safety checks

Code Quality Improvements:
- Reduced total lines by ~251 lines (-24% code reduction)
- Eliminated duplicate validation logic
- Improved TypeScript type safety
- Better component composition and separation of concerns
- Enhanced accessibility with proper ARIA labels

These changes make forms more intuitive, responsive, and easier to
maintain while reducing bundle size and improving runtime performance.

* style(ui): modernize component layouts and visual design

Update UI components with improved layouts, visual hierarchy, and
modern design patterns for better user experience.

Navigation & Brand Components:
- AppSwitcher
  * Enhanced visual design with better spacing
  * Improved active state indicators
  * Smoother transitions and hover effects
  * Better mobile responsiveness
- BrandIcons
  * Optimized icon rendering performance
  * Added support for more provider icons
  * Improved SVG handling and fallbacks
  * Better scaling across different screen sizes

Editor Components:
- JsonEditor
  * Enhanced syntax highlighting
  * Better error visualization
  * Improved code formatting options
  * Added line numbers and code folding support
- UsageScriptModal
  * Complete layout overhaul (1239 lines refactored)
  * Better script editor integration
  * Improved template selection UI
  * Enhanced preview and testing panels
  * Better error feedback and validation

Provider Components:
- ProviderCard
  * Redesigned card layout with modern aesthetics
  * Better information density and readability
  * Improved action buttons placement
  * Enhanced status indicators (active/inactive)
- ProviderList
  * Better grid/list view layouts
  * Improved drag-and-drop visual feedback
  * Enhanced sorting indicators
- ProviderActions
  * Streamlined action menu
  * Better icon consistency
  * Improved tooltips and accessibility

Usage & Footer:
- UsageFooter
  * Redesigned footer layout
  * Better quota visualization
  * Improved refresh controls
  * Enhanced error states

Design System Updates:
- dialog.tsx (shadcn/ui component)
  * Updated to latest design tokens
  * Better overlay animations
  * Improved focus management
- index.css
  * Added 65 lines of global utility classes
  * New animation keyframes
  * Enhanced color variables for dark mode
  * Improved typography scale
- tailwind.config.js
  * Extended theme with new design tokens
  * Added custom animations and transitions
  * New spacing and sizing utilities
  * Enhanced color palette

Visual Improvements:
- Consistent border radius across components
- Unified shadow system for depth perception
- Better color contrast for accessibility (WCAG AA)
- Smoother animations and transitions
- Improved dark mode support

These changes create a more polished, modern interface while
maintaining consistency with the application's design language.

* chore: update dialogs, i18n and improve component integration

Various functional updates and improvements across provider dialogs,
MCP panel, skills page, and internationalization.

Provider Dialogs:
- AddProviderDialog
  * Simplified form state management
  * Improved preset selection workflow
  * Better validation error messages
  * Enhanced template variable handling
- EditProviderDialog
  * Streamlined edit flow with better state synchronization
  * Improved handling of live config backfilling
  * Better error recovery for failed updates
  * Enhanced integration with parent components

MCP & Skills:
- UnifiedMcpPanel
  * Reduced complexity from 140+ to ~95 lines
  * Improved multi-app server management
  * Better server type detection (stdio/http)
  * Enhanced server status indicators
  * Cleaner integration with MCP form modal
- SkillsPage
  * Simplified navigation and state management
  * Better integration with RepoManagerPanel
  * Improved error handling for repository operations
  * Enhanced loading states
- SkillCard
  * Minor layout adjustments
  * Better action button placement

Environment & Configuration:
- EnvWarningBanner
  * Improved conflict detection messages
  * Better visual hierarchy for warnings
  * Enhanced dismissal behavior
- tauri.conf.json
  * Updated build configuration
  * Added new window management options

Internationalization:
- en.json & zh.json
  * Added 17 new translation keys for new features
  * Updated existing keys for better clarity
  * Added translations for new settings page
  * Improved consistency across UI text

Code Cleanup:
- mutations.ts
  * Removed 14 lines of unused mutation definitions
  * Cleaned up deprecated query invalidation logic
  * Better type safety for mutation parameters

Overall Impact:
- Reduced total lines by 51 (-10% in affected files)
- Improved component integration and data flow
- Better error handling and user feedback
- Enhanced i18n coverage for new features

These changes improve the overall polish and integration of various
components while removing technical debt and unused code.

* feat(backend): add auto-launch functionality

Implement system auto-launch feature to allow CC-Switch to start
automatically on system boot, improving user convenience.

Backend Implementation:
- auto_launch.rs: New module for auto-launch management
  * Cross-platform support using auto-launch crate
  * Enable/disable auto-launch with system integration
  * Proper error handling for permission issues
  * Platform-specific implementations (macOS/Windows/Linux)

Command Layer:
- Add get_auto_launch command to check current status
- Add set_auto_launch command to toggle auto-start
- Integrate commands with settings API

Settings Integration:
- Extend Settings struct with auto_launch field
- Persist auto-launch preference in settings store
- Automatic state synchronization on app startup

Dependencies:
- Add auto-launch ^0.5.0 to Cargo.toml
- Update Cargo.lock with new dependency tree

Technical Details:
- Uses platform-specific auto-launch mechanisms:
  * macOS: Login Items via LaunchServices
  * Windows: Registry Run key
  * Linux: XDG autostart desktop files
- Handles edge cases like permission denials gracefully
- Maintains settings consistency across app restarts

This feature enables users to have CC-Switch readily available
after system boot without manual intervention, particularly useful
for users who frequently switch between API providers.

* refactor(settings): enhance settings page with auto-launch integration

Complete refactoring of settings page architecture to integrate auto-launch
feature and improve overall settings management workflow.

SettingsPage Component:
- Integrate auto-launch toggle with WindowSettings section
- Improve layout and spacing for better visual hierarchy
- Enhanced error handling for settings operations
- Better loading states during settings updates
- Improved accessibility with proper ARIA labels

WindowSettings Component:
- Add auto-launch switch with real-time status
- Integrate with backend auto-launch commands
- Proper error feedback for permission issues
- Visual indicators for current auto-launch state
- Tooltip guidance for auto-launch functionality

useSettings Hook (Major Refactoring):
- Complete rewrite reducing complexity by ~30%
- Better separation of concerns with dedicated handlers
- Improved state management using React Query
- Enhanced auto-launch state synchronization
  * Fetch auto-launch status on mount
  * Real-time updates on toggle
  * Proper error recovery
- Optimized re-renders with better memoization
- Cleaner API for component integration
- Better TypeScript type safety

Settings API:
- Add getAutoLaunch() method
- Add setAutoLaunch(enabled: boolean) method
- Type-safe Tauri command invocations
- Proper error propagation to UI layer

Architecture Improvements:
- Reduced hook complexity from 197 to ~140 effective lines
- Eliminated redundant state management logic
- Better error boundaries and fallback handling
- Improved testability with clearer separation

User Experience Enhancements:
- Instant visual feedback on auto-launch toggle
- Clear error messages for permission issues
- Loading indicators during async operations
- Consistent behavior across all platforms

This refactoring provides a solid foundation for future settings
additions while maintaining code quality and user experience.

* refactor(ui): optimize FullScreenPanel, Dialog and App routing

Comprehensive refactoring of core UI components to improve code quality,
maintainability, and user experience.

FullScreenPanel Component:
- Enhanced props interface with better TypeScript types
- Improved layout flexibility with customizable padding
- Better header/footer composition patterns
- Enhanced scroll behavior for long content
- Added support for custom actions in header
- Improved responsive design for different screen sizes
- Better integration with parent components
- Cleaner prop drilling with context where appropriate

Dialog Component (shadcn/ui):
- Updated to latest component patterns
- Improved animation timing and easing
- Better focus trap management
- Enhanced overlay styling with backdrop blur
- Improved accessibility (ARIA labels, keyboard navigation)
- Better close button positioning and styling
- Enhanced mobile responsiveness
- Cleaner composition with DialogHeader/Footer

App Component Routing:
- Refactored routing logic for better clarity
- Improved state management for navigation
- Better integration with settings page
- Enhanced error boundary handling
- Cleaner separation of layout concerns
- Improved provider context propagation
- Better handling of deep links
- Optimized re-renders with React.memo where appropriate

Code Quality Improvements:
- Reduced prop drilling with better component composition
- Improved TypeScript type safety
- Better separation of concerns
- Enhanced code readability with clearer naming
- Eliminated redundant logic

Performance Optimizations:
- Reduced unnecessary re-renders
- Better memoization of callbacks
- Optimized component tree structure
- Improved event handler efficiency

User Experience:
- Smoother transitions and animations
- Better visual feedback for interactions
- Improved loading states
- More consistent behavior across features

These changes create a more maintainable and performant foundation
for the application's UI layer while improving the overall user
experience with smoother interactions and better visual polish.

* refactor(features): modernize Skills, Prompts and Agents components

Major refactoring of feature components to improve code quality,
user experience, and maintainability.

SkillsPage Component (299 lines refactored):
- Complete rewrite of layout and state management
- Better integration with RepoManagerPanel
- Improved navigation between list and detail views
- Enhanced error handling with user-friendly messages
- Better loading states with skeleton screens
- Optimized re-renders with proper memoization
- Cleaner separation between list and form views
- Improved skill card interactions
- Better responsive design for different screen sizes

RepoManagerPanel Component (370 lines refactored):
- Streamlined repository management workflow
- Enhanced form validation with real-time feedback
- Improved repository list with better visual hierarchy
- Better handling of git operations (clone, pull, delete)
- Enhanced error recovery for network issues
- Cleaner state management reducing complexity
- Improved TypeScript type safety
- Better integration with Skills backend API
- Enhanced loading indicators for async operations

PromptPanel Component (249 lines refactored):
- Modernized layout with FullScreenPanel integration
- Better separation between list and edit modes
- Improved prompt card design with better readability
- Enhanced search and filter functionality
- Cleaner state management for editing workflow
- Better integration with PromptFormPanel
- Improved delete confirmation with safety checks
- Enhanced keyboard navigation support

PromptFormPanel Component (238 lines refactored):
- Streamlined form layout and validation
- Better markdown editor integration
- Real-time preview with syntax highlighting
- Improved validation error display
- Enhanced save/cancel workflow
- Better handling of large prompt content
- Cleaner form state management
- Improved accessibility features

AgentsPanel Component (33 lines modified):
- Minor layout adjustments for consistency
- Better integration with FullScreenPanel
- Improved placeholder states
- Enhanced error boundaries

Type Definitions (types.ts):
- Added 10 new type definitions
- Better type safety for Skills/Prompts/Agents
- Enhanced interfaces for repository management
- Improved typing for form validations

Architecture Improvements:
- Reduced component coupling
- Better prop interfaces with explicit types
- Improved error boundaries
- Enhanced code reusability
- Better testing surface

User Experience Enhancements:
- Smoother transitions between views
- Better visual feedback for actions
- Improved error messages
- Enhanced loading states
- More intuitive navigation flows
- Better responsive layouts

Code Quality:
- Net reduction of 29 lines while adding features
- Improved code organization
- Better naming conventions
- Enhanced documentation
- Cleaner control flow

These changes significantly improve the maintainability and user
experience of core feature components while establishing consistent
patterns for future development.

* style(ui): refine component layouts and improve visual consistency

Comprehensive UI polish across multiple components to enhance visual
design, improve user experience, and maintain consistency.

UsageScriptModal Component (1302 lines refactored):
- Complete layout overhaul for better usability
- Improved script editor with syntax highlighting
- Better template selection interface
- Enhanced test/preview panels with clearer separation
- Improved error feedback and validation messages
- Better modal sizing and responsiveness
- Cleaner tab navigation between sections
- Enhanced code formatting and readability
- Improved loading states for async operations
- Better integration with parent components

MCP Components:
- McpFormModal (42 lines):
  * Streamlined form layout
  * Better server type selection (stdio/http)
  * Improved field grouping and labels
  * Enhanced validation feedback
- UnifiedMcpPanel (14 lines):
  * Minor layout adjustments
  * Better list item spacing
  * Improved server status indicators
  * Enhanced action button placement

Provider Components:
- ProviderCard (11 lines):
  * Refined card layout and spacing
  * Better visual hierarchy
  * Improved badge placement
  * Enhanced hover effects
- ProviderList (5 lines):
  * Minor grid layout adjustments
  * Better drag-and-drop visual feedback
- GeminiConfigSections (4 lines):
  * Field label alignment
  * Improved spacing consistency

Editor & Footer Components:
- JsonEditor (13 lines):
  * Better editor height management
  * Improved error display
  * Enhanced syntax highlighting
- UsageFooter (10 lines):
  * Refined footer layout
  * Better quota display
  * Improved refresh button placement

Settings & Environment:
- ImportExportSection (24 lines):
  * Better button layout
  * Improved action grouping
  * Enhanced visual feedback
- EnvWarningBanner (4 lines):
  * Refined alert styling
  * Better dismiss button placement

Global Styles (index.css):
- Added 11 lines of utility classes
- Improved transition timing
- Better focus indicators
- Enhanced scrollbar styling
- Refined spacing utilities

Design Improvements:
- Consistent spacing using design tokens
- Unified color palette application
- Better typography hierarchy
- Improved shadow system for depth
- Enhanced interactive states (hover, active, focus)
- Better border radius consistency
- Refined animation timings

Accessibility:
- Improved focus indicators
- Better keyboard navigation
- Enhanced screen reader support
- Improved color contrast ratios

Code Quality:
- Net increase of 68 lines due to UsageScriptModal improvements
- Better component organization
- Cleaner style application
- Reduced style duplication

These visual refinements create a more polished and professional
interface while maintaining excellent usability and accessibility
standards across all components.

* chore(i18n): add auto-launch translation keys

Add translation keys for new auto-launch feature to support
multi-language interface.

Translation Keys Added:
- autoLaunch: Label for auto-launch toggle
- autoLaunchDescription: Explanation of auto-launch functionality
- autoLaunchEnabled: Status message when enabled

Languages Updated:
- Chinese (zh.json): 简体中文翻译
- English (en.json): English translations

The translations maintain consistency with existing terminology
and provide clear, user-friendly descriptions of the auto-launch
feature across both supported languages.

* test: update test suites to match component refactoring

Comprehensive test updates to align with recent component refactoring
and new auto-launch functionality.

Component Tests:
- AddProviderDialog.test.tsx (10 lines):
  * Updated test cases for new dialog behavior
  * Enhanced mock data for preset selection
  * Improved assertions for validation

- ImportExportSection.test.tsx (16 lines):
  * Updated for new settings page integration
  * Enhanced test coverage for error scenarios
  * Better mock state management

- McpFormModal.test.tsx (60 lines):
  * Extensive updates for form refactoring
  * New test cases for multi-app selection
  * Enhanced validation testing
  * Better coverage of stdio/http server types

- ProviderList.test.tsx (11 lines):
  * Updated for new card layout
  * Enhanced drag-and-drop testing

- SettingsDialog.test.tsx (96 lines):
  * Major updates for SettingsPage migration
  * New test cases for auto-launch functionality
  * Enhanced integration test coverage
  * Better async operation testing

Hook Tests:
- useDirectorySettings.test.tsx (32 lines):
  * Updated for refactored hook logic
  * Enhanced test coverage for edge cases

- useDragSort.test.tsx (36 lines):
  * Simplified test cases
  * Better mock implementation
  * Improved assertions

- useImportExport tests (16 lines total):
  * Updated for new error handling
  * Enhanced test coverage

- useMcpValidation.test.tsx (23 lines):
  * Updated validation test cases
  * Better coverage of error scenarios

- useProviderActions.test.tsx (48 lines):
  * Extensive updates for hook refactoring
  * New test cases for provider operations
  * Enhanced mock data

- useSettings.test.tsx (12 lines):
  * New test cases for auto-launch
  * Enhanced settings state testing
  * Better async operation coverage

Integration Tests:
- App.test.tsx (41 lines):
  * Updated for new routing logic
  * Enhanced navigation testing
  * Better component integration coverage

- SettingsDialog.test.tsx (88 lines):
  * Complete rewrite for SettingsPage
  * New integration test scenarios
  * Enhanced user workflow testing

Mock Infrastructure:
- handlers.ts (117 lines):
  * Major updates for MSW handlers
  * New handlers for auto-launch commands
  * Enhanced error simulation
  * Better request/response mocking

- state.ts (37 lines):
  * Updated mock state structure
  * New state for auto-launch
  * Enhanced state reset functionality

- tauriMocks.ts (10 lines):
  * Updated mock implementations
  * Better type safety

- server.ts & testQueryClient.ts:
  * Minor cleanup (2 lines removed)

Test Infrastructure Improvements:
- Better test isolation
- Enhanced mock data consistency
- Improved async operation testing
- Better error scenario coverage
- Enhanced integration test patterns

Coverage Improvements:
- Net increase of 195 lines of test code
- Better coverage of edge cases
- Enhanced error path testing
- Improved integration test scenarios
- Better mock infrastructure

All tests now pass with the refactored components while maintaining
comprehensive coverage of functionality and edge cases.

* style(ui): improve window dragging and provider card styles

* fix(skills): resolve third-party skills installation failure

- Add skills_path field to Skill struct
- Use skills_path to construct correct source path during installation
- Fix installation for repos with custom skill subdirectories

* feat(icon): add icon type system and intelligent inference logic

Introduce a new icon system for provider customization:

- Add IconMetadata and IconPreset interfaces in src/types/icon.ts
  - Define structure for icon name, display name, category, keywords
  - Support default color configuration per icon

- Implement smart icon inference in src/config/iconInference.ts
  - Create iconMappings for 25+ AI providers and cloud platforms
  - Include Claude, DeepSeek, Qwen, Kimi, Google, AWS, Azure, etc.
  - inferIconForPreset(): match provider name to icon config
  - addIconsToPresets(): batch apply icons to preset arrays
  - Support fuzzy matching for flexible name recognition

This foundation enables automatic icon assignment when users add
providers, improving visual identification in the provider list.

* feat(ui): add icon picker, color picker and provider icon components

Implement comprehensive icon selection system for provider customization:

## New Components

### ProviderIcon (src/components/ProviderIcon.tsx)
- Render SVG icons by name with automatic fallback
- Display provider initials when icon not found
- Support custom sizing via size prop
- Use dangerouslySetInnerHTML for inline SVG rendering

### IconPicker (src/components/IconPicker.tsx)
- Grid-based icon selection with visual preview
- Real-time search filtering by name and keywords
- Integration with icon metadata for display names
- Responsive grid layout (6-10 columns based on screen)

### ColorPicker (src/components/ColorPicker.tsx)
- 12 preset colors for quick selection
- Native color input for custom color picking
- Hex input field for precise color entry
- Visual feedback for selected color state

## Icon Assets (src/icons/extracted/)
- 38 high-quality SVG icons for AI providers and platforms
- Includes: OpenAI, Claude, DeepSeek, Qwen, Kimi, Gemini, etc.
- Cloud platforms: AWS, Azure, Google Cloud, Cloudflare
- Auto-generated index.ts with getIcon/hasIcon helpers
- Metadata system with searchable keywords per icon

## Build Scripts
- scripts/extract-icons.js: Extract icons from simple-icons
- scripts/generate-icon-index.js: Generate TypeScript index file

* feat(provider): integrate icon system into provider UI components

Add icon customization support to provider management interface:

## Type System Updates

### Provider Interface (src/types.ts)
- Add optional `icon` field for icon name (e.g., "openai", "anthropic")
- Add optional `iconColor` field for hex color (e.g., "#00A67E")

### Form Schema (src/lib/schemas/provider.ts)
- Extend providerSchema with icon and iconColor optional fields
- Maintain backward compatibility with existing providers

## UI Components

### ProviderCard (src/components/providers/ProviderCard.tsx)
- Display ProviderIcon alongside provider name
- Add icon container with hover animation effect
- Adjust layout spacing for icon placement
- Update translate offsets for action buttons

### BasicFormFields (src/components/providers/forms/BasicFormFields.tsx)
- Add icon preview section showing current selection
- Implement fullscreen icon picker dialog
- Auto-apply default color from icon metadata on selection
- Display provider name and icon status in preview

### AddProviderDialog & EditProviderDialog
- Pass icon fields through form submission
- Preserve icon data during provider updates

This enables users to visually distinguish providers in the list
with custom icons, improving UX for multi-provider setups.

* feat(backend): add icon fields to Provider model and default mappings

Extend Rust backend to support provider icon customization:

## Provider Model (src-tauri/src/provider.rs)
- Add `icon: Option<String>` field for icon name
- Add `icon_color: Option<String>` field for hex color
- Use serde rename `iconColor` for frontend compatibility
- Apply skip_serializing_if for clean JSON output
- Update Provider::new() to initialize icon fields as None

## Provider Defaults (src-tauri/src/provider_defaults.rs) [NEW]
- Define ProviderIcon struct with name and color fields
- Create DEFAULT_PROVIDER_ICONS static HashMap with 23 providers:
  - AI providers: OpenAI, Anthropic, Claude, Google, Gemini,
    DeepSeek, Kimi, Moonshot, Zhipu, MiniMax, Baidu, Alibaba,
    Tencent, Meta, Microsoft, Cohere, Perplexity, Mistral, HuggingFace
  - Cloud platforms: AWS, Azure, Huawei, Cloudflare
- Implement infer_provider_icon() with exact and fuzzy matching
- Add unit tests for matching logic (exact, fuzzy, case-insensitive)

## Deep Link Support (src-tauri/src/deeplink.rs)
- Initialize icon fields when creating Provider from deep link import

## Module Registration (src-tauri/src/lib.rs)
- Register provider_defaults module

## Dependencies (Cargo.toml)
- Add once_cell for lazy static initialization

This backend support enables icon persistence and future features
like auto-icon inference during provider creation.

* chore(i18n): add translations for icon picker and provider icon

Add Chinese and English translations for icon customization feature:

## Icon Picker (iconPicker)
- search: "Search Icons" / "搜索图标"
- searchPlaceholder: "Enter icon name..." / "输入图标名称..."
- noResults: "No matching icons found" / "未找到匹配的图标"
- category.aiProvider: "AI Providers" / "AI 服务商"
- category.cloud: "Cloud Platforms" / "云平台"
- category.tool: "Dev Tools" / "开发工具"
- category.other: "Other" / "其他"

## Provider Icon (providerIcon)
- label: "Icon" / "图标"
- colorLabel: "Icon Color" / "图标颜色"
- selectIcon: "Select Icon" / "选择图标"
- preview: "Preview" / "预览"

These translations support the new icon picker UI components
and provider form icon selection interface.

* style(ui): refine header layout and AppSwitcher color scheme

Improve application header and component styling:

## App.tsx Header Layout
- Wrap title and settings button in flex container with gap
- Add vertical divider between title and settings icon
- Apply responsive padding (pl-1 sm:pl-2)
- Reformat JSX for better readability (prettier)
- Fix string template formatting in className

## AppSwitcher Color Update
- Change Claude tab gradient from orange/amber to teal/emerald/green
- Update shadow color to match new teal theme
- Change hover color from orange-500 to teal-500
- Align visual style with emerald/teal brand colors

## Dialog Component Cleanup
- Remove default close button (X icon) from DialogContent
- Allow parent components to control close button placement
- Remove unused lucide-react X import

## index.css Header Border
- Add top border (2px solid) to glass-header
- Apply to both light and dark mode variants
- Improve visual separation of header area

These changes enhance visual consistency and modernize the UI
appearance with a cohesive teal color scheme.

* chore(deps): add icon library and update preset configurations

Add dependencies and utility scripts for icon system:

## Dependencies (package.json)
- Add @lobehub/icons-static-svg@1.73.0
  - High-quality SVG icon library for AI providers
  - Source for extracted icons in src/icons/extracted/
- Update pnpm-lock.yaml accordingly

## Provider Preset Updates (src/config/claudeProviderPresets.ts)
- Add optional `icon` and `iconColor` fields to ProviderPreset interface
- Apply to Anthropic Official preset as example:
  - icon: "anthropic"
  - iconColor: "#D4915D"
- Future presets can include default icon configurations

## Utility Script (scripts/filter-icons.js) [NEW]
- Helper script for filtering and managing icon assets
- Supports icon discovery and validation workflow
- Complements extract-icons.js and generate-icon-index.js

This completes the icon system infrastructure, providing all
necessary tools and dependencies for icon customization.

* refactor(ui): simplify AppSwitcher styles and migrate to local SVG icons

- Replace complex gradient animations with clean, minimal tab design
- Migrate from @lobehub/icons CDN to local SVG assets for better reliability
- Fix clippy warning in error.rs (use inline format args)
- Improve code formatting in skill service and commands
- Reduce CSS complexity in AppSwitcher component (removed blur effects and gradients)
- Update BrandIcons to use imported local SVG files instead of dynamic image loading

This improves performance, reduces external dependencies, and provides a cleaner UI experience.

* style(ui): hide scrollbars across all browsers and optimize form layout

- Hide scrollbars globally with cross-browser support:
  * WebKit browsers (Chrome, Safari, Edge): ::-webkit-scrollbar { display: none }
  * Firefox: scrollbar-width: none
  * IE 10+: -ms-overflow-style: none
- Remove custom scrollbar styling (width, colors, hover states)
- Reorganize BasicFormFields layout:
  * Move icon picker to top center as a clickable preview (80x80)
  * Change name and notes fields to horizontal grid layout (md:grid-cols-2)
  * Remove icon preview section from bottom
  * Improve visual hierarchy and space efficiency

This provides a cleaner, more modern UI with hidden scrollbars while maintaining full scroll functionality.

* refactor(layout): standardize max-width to 60rem and optimize padding structure

- Unify container max-width across components:
  * Replace max-w-4xl with max-w-[60rem] in App.tsx provider list
  * Replace max-w-5xl with max-w-[60rem] in PromptPanel
  * Move padding from header element to inner container for consistent spacing
- Optimize padding hierarchy:
  * Remove px-6 from header element, add to inner header container
  * Remove px-6 from main element, keep it only in provider list container
  * Consolidate PromptPanel padding: move px-6 from nested divs to outer container
  * Remove redundant pl-1 sm:pl-2 from logo/title area
- Benefits:
  * Consistent 60rem max-width provides better readability on wide screens
  * Simplified padding structure reduces CSS complexity
  * Cleaner responsive behavior with unified spacing rules

This creates a more maintainable and visually consistent layout system.

* refactor(ui): unify layout system with 60rem max-width and consistent spacing

- Standardize container max-width across all panels:
  * Replace max-w-4xl and max-w-5xl with unified max-w-[60rem]
  * Apply to SettingsPage, UnifiedMcpPanel, SkillsPage, and FullScreenPanel
  * Ensures consistent reading width and visual balance on wide screens

- Optimize padding hierarchy and structure:
  * Move px-6 from parent elements to content containers
  * FullScreenPanel: Add max-w-[60rem] wrapper to header, content, and footer
  * Add border separators (border-t/border-b) to header and footer sections
  * Consolidate nested padding in MCP, Skills, and Prompts panels
  * Remove redundant padding layers for cleaner CSS

- Simplify component styling:
  * MCP list items: Replace card-based layout with modern group-based design
  * Remove unnecessary wrapper divs and flatten DOM structure
  * Update card hover effects with smooth transitions
  * Simplify icon selection dialog (remove description text in BasicFormFields)

- Benefits:
  * Consistent 60rem max-width provides optimal readability
  * Unified spacing rules reduce maintenance complexity
  * Cleaner component hierarchy improves performance
  * Better responsive behavior across different screen sizes
  * More cohesive visual design language throughout the app

This creates a maintainable, scalable design system foundation.

* feat(deeplink): add Claude model fields support and enhance import dialog

- Add Claude-specific model field support in deeplink import:
  * Support model (ANTHROPIC_MODEL) - general default model
  * Support haikuModel (ANTHROPIC_DEFAULT_HAIKU_MODEL)
  * Support sonnetModel (ANTHROPIC_DEFAULT_SONNET_MODEL)
  * Support opusModel (ANTHROPIC_DEFAULT_OPUS_MODEL)
  * Backend: Update DeepLinkImportRequest struct to include optional model fields
  * Frontend: Add TypeScript type definitions for new model parameters

- Enhance deeplink demo page (deplink.html):
  * Add 5 new Claude configuration examples showcasing different model setups
  * Add parameter documentation with required/optional tags
  * Include basic config (no models), single model, complete 4-model, partial models, and third-party provider examples
  * Improve visual design with param-list component and color-coded badges
  * Add detailed descriptions for each configuration scenario

- Redesign DeepLinkImportDialog layout:
  * Switch from 3-column to compact 2-column grid layout
  * Reduce dialog width from 500px to 650px for better content display
  * Add dedicated section for Claude model configurations with blue highlight box
  * Use uppercase labels and smaller text for more information density
  * Add truncation and tooltips for long URLs
  * Improve visual hierarchy with spacing and grouping
  * Increase z-index to 9999 to ensure dialog appears on top

- Minor UI refinements:
  * Update App.tsx layout adjustments
  * Optimize McpFormModal styling
  * Refine ProviderCard and BasicFormFields components

This enables users to import Claude providers with precise model configurations via deeplink.

* feat(deeplink): add config file support for deeplink import

Support importing provider configuration from embedded or remote config files.
- Add base64 dependency for config content encoding
- Support config, configFormat, and configUrl parameters
- Make homepage/endpoint/apiKey optional when config is provided
- Add config parsing and merging logic

* feat(deeplink): enhance dialog with config file preview

Add config file parsing and preview in deep link import dialog.
- Support Base64 encoded config display
- Add config file source indicator (embedded/remote)
- Parse and display config fields by app type (Claude/Codex/Gemini)
- Mask sensitive values in config preview
- Improve dialog layout and content organization

* refactor(ui): unify dialog styles and improve layout consistency

Standardize dialog and panel components across the application.
- Update dialog background to use semantic color tokens
- Adjust FullScreenPanel max-width to 56rem for better alignment
- Add drag region and prevent body scroll in full-screen panels
- Optimize button sizes and spacing in panel headers
- Apply consistent styling to all dialog-based components

* i18n: add deeplink config preview translations

Add missing translation keys for config file preview feature.
- Add configSource, configEmbedded, configRemote labels
- Add configDetails and configUrl display strings
- Support both Chinese and English versions

* feat(deeplink): enhance test page with v3.8 config file examples

Improve deeplink test page with comprehensive config file import examples.
- Add version badge for v3.8 features
- Add copy-to-clipboard functionality for all deep links
- Add Claude config file import examples (embedded/remote)
- Add Codex config file import examples (auth.json + config.toml)
- Add Gemini config file import examples (.env format)
- Add config generator tool for easy testing
- Update UI with better styling and layout

* feat(settings): add autoSaveSettings for lightweight auto-save

Add optimized auto-save function for General tab settings.
- Add autoSaveSettings method for non-destructive auto-save
- Only trigger system APIs when values actually change
- Avoid unnecessary auto-launch and plugin config updates
- Update tests to cover new functionality

* refactor(settings): simplify settings page layout and auto-save

Reorganize settings page structure and integrate autoSaveSettings.
- Move save button inline within Advanced tab content
- Remove sticky footer for cleaner layout
- Use autoSaveSettings for General tab settings
- Simplify dialog close behavior
- Refactor ImportExportSection layout

* style(providers): optimize card layout and action button sizes

Improve provider card visual density and action buttons.
- Reduce icon button sizes for compact layout
- Adjust drag handle and icon sizes
- Tighten spacing between action buttons
- Update hover translate values for better alignment

* refactor(mcp): improve form modal layout with adaptive height editor

Restructure MCP form modal for better space utilization.
- Split form into upper form fields and lower JSON editor sections
- Add full-height mode support for JsonEditor component
- Use flex layout for editor to fill available space
- Update PromptFormPanel to use full-height editor
- Fix locale text formatting

* style: unify list item styles with semantic colors

Apply consistent styling to list items across components.
- Replace hardcoded colors with semantic tokens in MCP and Prompt list items
- Add glass effect container to EndpointSpeedTest panel
- Format code for better readability

* style: format template literals for better readability

Improve code formatting for conditional className expressions.
- Break long template literals across multiple lines
- Maintain consistent formatting in MCP form and endpoint test components

* feat(deeplink): add config merge command for preview

Expose config merging functionality to frontend for preview.
- Add merge_deeplink_config Tauri command
- Make parse_and_merge_config public for reuse
- Enable frontend to display complete config before import

* feat(deeplink): merge and display config in import dialog

Enhance import dialog to fetch and display complete config.
- Call mergeDeeplinkConfig API when config is present
- Add UTF-8 base64 decoding support for config content
- Add scrollable content area with custom scrollbar styling
- Show complete configuration before user confirms import

* i18n: add config merge error message

Add translation for config file merge error handling.

* style(deeplink): format test page HTML for better readability

Improve HTML formatting in deeplink test page.
- Format multiline attributes for better readability
- Add consistent indentation to nested elements
- Break long lines in buttons and links

* refactor(usage): improve footer layout with two-row design

Reorganize usage footer for better readability and space efficiency.
- Split into two rows: update time + refresh button (row 1), usage stats (row 2)
- Move refresh button to top row next to update time
- Remove card background for cleaner look
- Add fallback text when never updated
- Improve spacing and alignment
- Format template literals for consistency

* feat(database): add SQLite database infrastructure

- Add rusqlite dependency (v0.32.1) and r2d2 connection pooling
- Implement Database module with CRUD operations for providers, MCP servers, prompts, and skills
- Add schema initialization with proper indexes
- Include data migration utilities from JSON config to SQLite
- Support timestamp tracking (created_at, updated_at)

* refactor(core): integrate SQLite database into application core

- Initialize database on app startup with migration from JSON config
- Update AppState to include Database instance alongside MultiAppConfig
- Simplify store module by removing unused session management code
- Add database initialization to app setup flow
- Support both database and legacy config during transition

* refactor(services): migrate service layer to use SQLite database

- Refactor ProviderService to use database queries instead of in-memory config
- Update McpService to fetch and store MCP servers in database
- Migrate PromptService to database-backed storage
- Simplify ConfigService by removing complex transaction logic
- Remove 648 lines of redundant code through database abstraction

* refactor(commands): update command layer to use database API

- Update config commands to query database for providers and settings
- Modify provider commands to pass database handle to services
- Update MCP commands to use database-backed operations
- Refactor prompt and skill commands to leverage database storage
- Simplify import/export commands with database integration

* refactor(backend): update supporting modules for database compatibility

- Add DatabaseError variant to AppError enum
- Update provider module to support database-backed operations
- Modify codex_config to work with new database structure
- Ensure error handling covers database operations

* refactor(frontend): update UI components for database migration

- Update UsageFooter component to handle new data structure
- Modify SkillsPage to work with database-backed skills management
- Ensure frontend compatibility with refactored backend

* feat(skills): add search functionality to Skills page

- Add search input with Search icon in SkillsPage component
- Implement useMemo-based filtering by skill name, description, and directory
- Display search results count when filtering is active
- Show "no results" message when no skills match the search query
- Add i18n translations for search UI (zh/en)
- Maintain responsive layout and consistent styling with existing UI

* refactor(backend): replace unsafe unwrap calls with proper error handling

- Add to_json_string helper for safe JSON serialization
- Add lock_conn macro for safe Mutex locking
- Replace 41 unwrap() calls with proper error handling:
  - database.rs: JSON serialization and Mutex operations (31 fixes)
  - lib.rs: macOS NSWindow and tray icon handling (3 fixes)
  - services/provider.rs: Claude model normalization (1 fix)
  - services/prompt.rs: timestamp generation (3 fixes)
  - services/skill.rs: directory name extraction (2 fixes)
  - mcp.rs: HashMap initialization and type conversions (5 fixes)
  - app_config.rs: timestamp fallback (1 fix)

This improves application stability and prevents potential panics.

* feat(init): implement automatic data import on first launch

Add comprehensive first-launch data import system:

Database layer:
- Add is_empty_for_first_import() to detect empty database
- Add init_default_skill_repos() to initialize 3 default skill repositories

Services layer:
- Implement McpService::import_from_claude/codex/gemini()
  to import MCP servers from existing config files
- Implement PromptService::import_from_file_on_first_launch()
  to import prompt files (CLAUDE.md, AGENTS.md, GEMINI.md)

Startup flow (lib.rs):
- Check if database is empty on startup
- Import existing configurations if detected:
  1. Initialize default skill repositories
  2. Import provider configurations from live settings
  3. Import MCP servers from config files
  4. Import prompt files
- All imports are fault-tolerant and logged

This ensures seamless migration from file-based configs to database.

* fix(skills): auto-sync locally installed skills to database

Add automatic synchronization in get_skills command:
- Detect locally installed skills in ~/.claude/skills/
- Auto-add to database if not already tracked
- Ensures existing skills are recognized on first launch

This fixes the issue where user's existing skills were not
imported into the database on initial application run.

* docs(frontend): add code comments and improve formatting

- Add explanatory comment in EditProviderDialog about config assembly
- Improve import formatting in SkillsPage for better readability

* feat(deeplink): display all four Claude model fields in import dialog

- Show haiku/sonnet/opus/multiModel fields conditionally for Claude
- Maintain single model field display for Codex and Gemini
- Add i18n translations for new model field labels (zh/en)

* feat(backend): add database SQL export/import with backup

- Enable rusqlite backup feature for SQL dump support
- Implement export_sql to generate SQLite-compatible SQL dumps
- Implement import_sql with automatic backup before import
- Add snapshot_to_memory to avoid long-held database locks
- Add backup rotation to retain latest 10 backups
- Support atomic import with rollback on failure

* refactor(backend): migrate import/export to use SQL backup

- Reimplement export_config_to_file to use database.export_sql
- Reimplement import_config_from_file to use database.import_sql
- Add sync_current_from_db to sync live configs after import
- Add settings database binding on app initialization
- Remove deprecated JSON-based config import logic

* refactor(backend): migrate settings storage to database

- Add bind_db function to initialize database-backed settings
- Implement load_initial_settings with database fallback
- Replace direct file save with settings store management
- Add SETTINGS_DB static for database binding
- Maintain backward compatibility with file-based settings
- Keep settings.json for legacy migration support

* feat(frontend): update import/export UI for SQL backup

- Change default export filename from .json to .sql
- Update file format to timestamp format (YYYYMMDD_HHMMSS)
- Update error messages to reference SQL files
- Align with backend SQL export/import implementation

* feat(i18n): update translations for SQL backup feature

- Update Chinese translations for SQL import/export
- Update English translations for SQL import/export
- Change terminology from 'config file' to 'SQL backup'
- Update error messages and UI hints

* fix(backend): remove unnecessary dereference in backup operation

- Simplify Backup::new call by removing redundant dereference
- MutexGuard already implements DerefMut
2025-11-24 11:00:45 +08:00
Jason cc0b9352bc update readme 2025-11-22 21:51:09 +08:00
Bill ZHANG 01d8bb53ac fix(codex): use http_headers instead of headers in MCP config (#276)
Codex CLI expects the field name to be `http_headers` (not `headers`)
in the MCP server configuration TOML format.

Changes:
- Update import_from_codex() to read from both `http_headers` (correct)
  and `headers` (legacy) with priority to `http_headers`
- Update json_server_to_toml_table() to write `http_headers` instead
  of `headers`
- Update core_fields lists to use `http_headers` for HTTP/SSE types

This follows the same pattern as the recent Gemini fix and ensures
backward compatibility while generating correct Codex-compatible configs.
2025-11-22 20:46:30 +08:00
YoVinchen d38fcd63ea Refactor/UI (#273)
* feat(components): add reusable full-screen panel components

Add new full-screen panel components to support the UI refactoring:

- FullScreenPanel: Reusable full-screen layout component with header,
  content area, and optional footer. Provides consistent layout for
  settings, prompts, and other full-screen views.

- PromptFormPanel: Dedicated panel for creating and editing prompts
  with markdown preview support. Features real-time validation and
  integrated save/cancel actions.

- AgentsPanel: Panel component for managing agent configurations.
  Provides a consistent interface for agent CRUD operations.

- RepoManagerPanel: Full-featured repository manager panel for Skills.
  Supports repository listing, addition, deletion, and configuration
  management with integrated validation.

These components establish the foundation for the upcoming settings
page migration from dialog-based to full-screen layout.

* refactor(settings): migrate from dialog to full-screen page layout

Complete migration of settings from modal dialog to dedicated full-screen
page, improving UX and providing more space for configuration options.

Changes:
- Remove SettingsDialog component (legacy modal-based interface)
- Add SettingsPage component with full-screen layout using FullScreenPanel
- Refactor App.tsx routing to support dedicated settings page
  * Add settings route handler
  * Update navigation logic from dialog-based to page-based
  * Integrate with existing app switcher and provider management
- Update ImportExportSection to work with new page layout
  * Improve spacing and layout for better readability
  * Enhanced error handling and user feedback
  * Better integration with page-level actions
- Enhance useSettings hook to support page-based workflow
  * Add navigation state management
  * Improve settings persistence logic
  * Better error boundary handling

Benefits:
- More intuitive navigation with dedicated settings page
- Better use of screen space for complex configurations
- Improved accessibility with clearer visual hierarchy
- Consistent with modern desktop application patterns
- Easier to extend with new settings sections

This change is part of the larger UI refactoring initiative to modernize
the application interface and improve user experience.

* refactor(forms): simplify and modernize form components

Comprehensive refactoring of form components to reduce complexity,
improve maintainability, and enhance user experience.

Provider Forms:
- CodexCommonConfigModal & CodexConfigSections
  * Simplified state management with reduced boilerplate
  * Improved field validation and error handling
  * Better layout with consistent spacing
  * Enhanced model selection with visual indicators
- GeminiCommonConfigModal & GeminiConfigSections
  * Streamlined authentication flow (OAuth vs API Key)
  * Cleaner form layout with better grouping
  * Improved validation feedback
  * Better integration with parent components
- CommonConfigEditor
  * Reduced from 178 to 68 lines (-62% complexity)
  * Extracted reusable form patterns
  * Improved JSON editing with syntax validation
  * Better error messages and recovery options
- EndpointSpeedTest
  * Complete rewrite for better UX
  * Real-time testing progress indicators
  * Enhanced error handling with retry logic
  * Visual feedback for test results (color-coded latency)

MCP & Prompts:
- McpFormModal
  * Simplified from 581 to ~360 lines
  * Better stdio/http server type handling
  * Improved form validation
  * Enhanced multi-app selection (Claude/Codex/Gemini)
- PromptPanel
  * Cleaner integration with PromptFormPanel
  * Improved list/grid view switching
  * Better state management for editing workflows
  * Enhanced delete confirmation with safety checks

Code Quality Improvements:
- Reduced total lines by ~251 lines (-24% code reduction)
- Eliminated duplicate validation logic
- Improved TypeScript type safety
- Better component composition and separation of concerns
- Enhanced accessibility with proper ARIA labels

These changes make forms more intuitive, responsive, and easier to
maintain while reducing bundle size and improving runtime performance.

* style(ui): modernize component layouts and visual design

Update UI components with improved layouts, visual hierarchy, and
modern design patterns for better user experience.

Navigation & Brand Components:
- AppSwitcher
  * Enhanced visual design with better spacing
  * Improved active state indicators
  * Smoother transitions and hover effects
  * Better mobile responsiveness
- BrandIcons
  * Optimized icon rendering performance
  * Added support for more provider icons
  * Improved SVG handling and fallbacks
  * Better scaling across different screen sizes

Editor Components:
- JsonEditor
  * Enhanced syntax highlighting
  * Better error visualization
  * Improved code formatting options
  * Added line numbers and code folding support
- UsageScriptModal
  * Complete layout overhaul (1239 lines refactored)
  * Better script editor integration
  * Improved template selection UI
  * Enhanced preview and testing panels
  * Better error feedback and validation

Provider Components:
- ProviderCard
  * Redesigned card layout with modern aesthetics
  * Better information density and readability
  * Improved action buttons placement
  * Enhanced status indicators (active/inactive)
- ProviderList
  * Better grid/list view layouts
  * Improved drag-and-drop visual feedback
  * Enhanced sorting indicators
- ProviderActions
  * Streamlined action menu
  * Better icon consistency
  * Improved tooltips and accessibility

Usage & Footer:
- UsageFooter
  * Redesigned footer layout
  * Better quota visualization
  * Improved refresh controls
  * Enhanced error states

Design System Updates:
- dialog.tsx (shadcn/ui component)
  * Updated to latest design tokens
  * Better overlay animations
  * Improved focus management
- index.css
  * Added 65 lines of global utility classes
  * New animation keyframes
  * Enhanced color variables for dark mode
  * Improved typography scale
- tailwind.config.js
  * Extended theme with new design tokens
  * Added custom animations and transitions
  * New spacing and sizing utilities
  * Enhanced color palette

Visual Improvements:
- Consistent border radius across components
- Unified shadow system for depth perception
- Better color contrast for accessibility (WCAG AA)
- Smoother animations and transitions
- Improved dark mode support

These changes create a more polished, modern interface while
maintaining consistency with the application's design language.

* chore: update dialogs, i18n and improve component integration

Various functional updates and improvements across provider dialogs,
MCP panel, skills page, and internationalization.

Provider Dialogs:
- AddProviderDialog
  * Simplified form state management
  * Improved preset selection workflow
  * Better validation error messages
  * Enhanced template variable handling
- EditProviderDialog
  * Streamlined edit flow with better state synchronization
  * Improved handling of live config backfilling
  * Better error recovery for failed updates
  * Enhanced integration with parent components

MCP & Skills:
- UnifiedMcpPanel
  * Reduced complexity from 140+ to ~95 lines
  * Improved multi-app server management
  * Better server type detection (stdio/http)
  * Enhanced server status indicators
  * Cleaner integration with MCP form modal
- SkillsPage
  * Simplified navigation and state management
  * Better integration with RepoManagerPanel
  * Improved error handling for repository operations
  * Enhanced loading states
- SkillCard
  * Minor layout adjustments
  * Better action button placement

Environment & Configuration:
- EnvWarningBanner
  * Improved conflict detection messages
  * Better visual hierarchy for warnings
  * Enhanced dismissal behavior
- tauri.conf.json
  * Updated build configuration
  * Added new window management options

Internationalization:
- en.json & zh.json
  * Added 17 new translation keys for new features
  * Updated existing keys for better clarity
  * Added translations for new settings page
  * Improved consistency across UI text

Code Cleanup:
- mutations.ts
  * Removed 14 lines of unused mutation definitions
  * Cleaned up deprecated query invalidation logic
  * Better type safety for mutation parameters

Overall Impact:
- Reduced total lines by 51 (-10% in affected files)
- Improved component integration and data flow
- Better error handling and user feedback
- Enhanced i18n coverage for new features

These changes improve the overall polish and integration of various
components while removing technical debt and unused code.

* feat(backend): add auto-launch functionality

Implement system auto-launch feature to allow CC-Switch to start
automatically on system boot, improving user convenience.

Backend Implementation:
- auto_launch.rs: New module for auto-launch management
  * Cross-platform support using auto-launch crate
  * Enable/disable auto-launch with system integration
  * Proper error handling for permission issues
  * Platform-specific implementations (macOS/Windows/Linux)

Command Layer:
- Add get_auto_launch command to check current status
- Add set_auto_launch command to toggle auto-start
- Integrate commands with settings API

Settings Integration:
- Extend Settings struct with auto_launch field
- Persist auto-launch preference in settings store
- Automatic state synchronization on app startup

Dependencies:
- Add auto-launch ^0.5.0 to Cargo.toml
- Update Cargo.lock with new dependency tree

Technical Details:
- Uses platform-specific auto-launch mechanisms:
  * macOS: Login Items via LaunchServices
  * Windows: Registry Run key
  * Linux: XDG autostart desktop files
- Handles edge cases like permission denials gracefully
- Maintains settings consistency across app restarts

This feature enables users to have CC-Switch readily available
after system boot without manual intervention, particularly useful
for users who frequently switch between API providers.

* refactor(settings): enhance settings page with auto-launch integration

Complete refactoring of settings page architecture to integrate auto-launch
feature and improve overall settings management workflow.

SettingsPage Component:
- Integrate auto-launch toggle with WindowSettings section
- Improve layout and spacing for better visual hierarchy
- Enhanced error handling for settings operations
- Better loading states during settings updates
- Improved accessibility with proper ARIA labels

WindowSettings Component:
- Add auto-launch switch with real-time status
- Integrate with backend auto-launch commands
- Proper error feedback for permission issues
- Visual indicators for current auto-launch state
- Tooltip guidance for auto-launch functionality

useSettings Hook (Major Refactoring):
- Complete rewrite reducing complexity by ~30%
- Better separation of concerns with dedicated handlers
- Improved state management using React Query
- Enhanced auto-launch state synchronization
  * Fetch auto-launch status on mount
  * Real-time updates on toggle
  * Proper error recovery
- Optimized re-renders with better memoization
- Cleaner API for component integration
- Better TypeScript type safety

Settings API:
- Add getAutoLaunch() method
- Add setAutoLaunch(enabled: boolean) method
- Type-safe Tauri command invocations
- Proper error propagation to UI layer

Architecture Improvements:
- Reduced hook complexity from 197 to ~140 effective lines
- Eliminated redundant state management logic
- Better error boundaries and fallback handling
- Improved testability with clearer separation

User Experience Enhancements:
- Instant visual feedback on auto-launch toggle
- Clear error messages for permission issues
- Loading indicators during async operations
- Consistent behavior across all platforms

This refactoring provides a solid foundation for future settings
additions while maintaining code quality and user experience.

* refactor(ui): optimize FullScreenPanel, Dialog and App routing

Comprehensive refactoring of core UI components to improve code quality,
maintainability, and user experience.

FullScreenPanel Component:
- Enhanced props interface with better TypeScript types
- Improved layout flexibility with customizable padding
- Better header/footer composition patterns
- Enhanced scroll behavior for long content
- Added support for custom actions in header
- Improved responsive design for different screen sizes
- Better integration with parent components
- Cleaner prop drilling with context where appropriate

Dialog Component (shadcn/ui):
- Updated to latest component patterns
- Improved animation timing and easing
- Better focus trap management
- Enhanced overlay styling with backdrop blur
- Improved accessibility (ARIA labels, keyboard navigation)
- Better close button positioning and styling
- Enhanced mobile responsiveness
- Cleaner composition with DialogHeader/Footer

App Component Routing:
- Refactored routing logic for better clarity
- Improved state management for navigation
- Better integration with settings page
- Enhanced error boundary handling
- Cleaner separation of layout concerns
- Improved provider context propagation
- Better handling of deep links
- Optimized re-renders with React.memo where appropriate

Code Quality Improvements:
- Reduced prop drilling with better component composition
- Improved TypeScript type safety
- Better separation of concerns
- Enhanced code readability with clearer naming
- Eliminated redundant logic

Performance Optimizations:
- Reduced unnecessary re-renders
- Better memoization of callbacks
- Optimized component tree structure
- Improved event handler efficiency

User Experience:
- Smoother transitions and animations
- Better visual feedback for interactions
- Improved loading states
- More consistent behavior across features

These changes create a more maintainable and performant foundation
for the application's UI layer while improving the overall user
experience with smoother interactions and better visual polish.

* refactor(features): modernize Skills, Prompts and Agents components

Major refactoring of feature components to improve code quality,
user experience, and maintainability.

SkillsPage Component (299 lines refactored):
- Complete rewrite of layout and state management
- Better integration with RepoManagerPanel
- Improved navigation between list and detail views
- Enhanced error handling with user-friendly messages
- Better loading states with skeleton screens
- Optimized re-renders with proper memoization
- Cleaner separation between list and form views
- Improved skill card interactions
- Better responsive design for different screen sizes

RepoManagerPanel Component (370 lines refactored):
- Streamlined repository management workflow
- Enhanced form validation with real-time feedback
- Improved repository list with better visual hierarchy
- Better handling of git operations (clone, pull, delete)
- Enhanced error recovery for network issues
- Cleaner state management reducing complexity
- Improved TypeScript type safety
- Better integration with Skills backend API
- Enhanced loading indicators for async operations

PromptPanel Component (249 lines refactored):
- Modernized layout with FullScreenPanel integration
- Better separation between list and edit modes
- Improved prompt card design with better readability
- Enhanced search and filter functionality
- Cleaner state management for editing workflow
- Better integration with PromptFormPanel
- Improved delete confirmation with safety checks
- Enhanced keyboard navigation support

PromptFormPanel Component (238 lines refactored):
- Streamlined form layout and validation
- Better markdown editor integration
- Real-time preview with syntax highlighting
- Improved validation error display
- Enhanced save/cancel workflow
- Better handling of large prompt content
- Cleaner form state management
- Improved accessibility features

AgentsPanel Component (33 lines modified):
- Minor layout adjustments for consistency
- Better integration with FullScreenPanel
- Improved placeholder states
- Enhanced error boundaries

Type Definitions (types.ts):
- Added 10 new type definitions
- Better type safety for Skills/Prompts/Agents
- Enhanced interfaces for repository management
- Improved typing for form validations

Architecture Improvements:
- Reduced component coupling
- Better prop interfaces with explicit types
- Improved error boundaries
- Enhanced code reusability
- Better testing surface

User Experience Enhancements:
- Smoother transitions between views
- Better visual feedback for actions
- Improved error messages
- Enhanced loading states
- More intuitive navigation flows
- Better responsive layouts

Code Quality:
- Net reduction of 29 lines while adding features
- Improved code organization
- Better naming conventions
- Enhanced documentation
- Cleaner control flow

These changes significantly improve the maintainability and user
experience of core feature components while establishing consistent
patterns for future development.

* style(ui): refine component layouts and improve visual consistency

Comprehensive UI polish across multiple components to enhance visual
design, improve user experience, and maintain consistency.

UsageScriptModal Component (1302 lines refactored):
- Complete layout overhaul for better usability
- Improved script editor with syntax highlighting
- Better template selection interface
- Enhanced test/preview panels with clearer separation
- Improved error feedback and validation messages
- Better modal sizing and responsiveness
- Cleaner tab navigation between sections
- Enhanced code formatting and readability
- Improved loading states for async operations
- Better integration with parent components

MCP Components:
- McpFormModal (42 lines):
  * Streamlined form layout
  * Better server type selection (stdio/http)
  * Improved field grouping and labels
  * Enhanced validation feedback
- UnifiedMcpPanel (14 lines):
  * Minor layout adjustments
  * Better list item spacing
  * Improved server status indicators
  * Enhanced action button placement

Provider Components:
- ProviderCard (11 lines):
  * Refined card layout and spacing
  * Better visual hierarchy
  * Improved badge placement
  * Enhanced hover effects
- ProviderList (5 lines):
  * Minor grid layout adjustments
  * Better drag-and-drop visual feedback
- GeminiConfigSections (4 lines):
  * Field label alignment
  * Improved spacing consistency

Editor & Footer Components:
- JsonEditor (13 lines):
  * Better editor height management
  * Improved error display
  * Enhanced syntax highlighting
- UsageFooter (10 lines):
  * Refined footer layout
  * Better quota display
  * Improved refresh button placement

Settings & Environment:
- ImportExportSection (24 lines):
  * Better button layout
  * Improved action grouping
  * Enhanced visual feedback
- EnvWarningBanner (4 lines):
  * Refined alert styling
  * Better dismiss button placement

Global Styles (index.css):
- Added 11 lines of utility classes
- Improved transition timing
- Better focus indicators
- Enhanced scrollbar styling
- Refined spacing utilities

Design Improvements:
- Consistent spacing using design tokens
- Unified color palette application
- Better typography hierarchy
- Improved shadow system for depth
- Enhanced interactive states (hover, active, focus)
- Better border radius consistency
- Refined animation timings

Accessibility:
- Improved focus indicators
- Better keyboard navigation
- Enhanced screen reader support
- Improved color contrast ratios

Code Quality:
- Net increase of 68 lines due to UsageScriptModal improvements
- Better component organization
- Cleaner style application
- Reduced style duplication

These visual refinements create a more polished and professional
interface while maintaining excellent usability and accessibility
standards across all components.

* chore(i18n): add auto-launch translation keys

Add translation keys for new auto-launch feature to support
multi-language interface.

Translation Keys Added:
- autoLaunch: Label for auto-launch toggle
- autoLaunchDescription: Explanation of auto-launch functionality
- autoLaunchEnabled: Status message when enabled

Languages Updated:
- Chinese (zh.json): 简体中文翻译
- English (en.json): English translations

The translations maintain consistency with existing terminology
and provide clear, user-friendly descriptions of the auto-launch
feature across both supported languages.

* test: update test suites to match component refactoring

Comprehensive test updates to align with recent component refactoring
and new auto-launch functionality.

Component Tests:
- AddProviderDialog.test.tsx (10 lines):
  * Updated test cases for new dialog behavior
  * Enhanced mock data for preset selection
  * Improved assertions for validation

- ImportExportSection.test.tsx (16 lines):
  * Updated for new settings page integration
  * Enhanced test coverage for error scenarios
  * Better mock state management

- McpFormModal.test.tsx (60 lines):
  * Extensive updates for form refactoring
  * New test cases for multi-app selection
  * Enhanced validation testing
  * Better coverage of stdio/http server types

- ProviderList.test.tsx (11 lines):
  * Updated for new card layout
  * Enhanced drag-and-drop testing

- SettingsDialog.test.tsx (96 lines):
  * Major updates for SettingsPage migration
  * New test cases for auto-launch functionality
  * Enhanced integration test coverage
  * Better async operation testing

Hook Tests:
- useDirectorySettings.test.tsx (32 lines):
  * Updated for refactored hook logic
  * Enhanced test coverage for edge cases

- useDragSort.test.tsx (36 lines):
  * Simplified test cases
  * Better mock implementation
  * Improved assertions

- useImportExport tests (16 lines total):
  * Updated for new error handling
  * Enhanced test coverage

- useMcpValidation.test.tsx (23 lines):
  * Updated validation test cases
  * Better coverage of error scenarios

- useProviderActions.test.tsx (48 lines):
  * Extensive updates for hook refactoring
  * New test cases for provider operations
  * Enhanced mock data

- useSettings.test.tsx (12 lines):
  * New test cases for auto-launch
  * Enhanced settings state testing
  * Better async operation coverage

Integration Tests:
- App.test.tsx (41 lines):
  * Updated for new routing logic
  * Enhanced navigation testing
  * Better component integration coverage

- SettingsDialog.test.tsx (88 lines):
  * Complete rewrite for SettingsPage
  * New integration test scenarios
  * Enhanced user workflow testing

Mock Infrastructure:
- handlers.ts (117 lines):
  * Major updates for MSW handlers
  * New handlers for auto-launch commands
  * Enhanced error simulation
  * Better request/response mocking

- state.ts (37 lines):
  * Updated mock state structure
  * New state for auto-launch
  * Enhanced state reset functionality

- tauriMocks.ts (10 lines):
  * Updated mock implementations
  * Better type safety

- server.ts & testQueryClient.ts:
  * Minor cleanup (2 lines removed)

Test Infrastructure Improvements:
- Better test isolation
- Enhanced mock data consistency
- Improved async operation testing
- Better error scenario coverage
- Enhanced integration test patterns

Coverage Improvements:
- Net increase of 195 lines of test code
- Better coverage of edge cases
- Enhanced error path testing
- Improved integration test scenarios
- Better mock infrastructure

All tests now pass with the refactored components while maintaining
comprehensive coverage of functionality and edge cases.

* style(ui): improve window dragging and provider card styles

* fix(skills): resolve third-party skills installation failure

- Add skills_path field to Skill struct
- Use skills_path to construct correct source path during installation
- Fix installation for repos with custom skill subdirectories

* feat(icon): add icon type system and intelligent inference logic

Introduce a new icon system for provider customization:

- Add IconMetadata and IconPreset interfaces in src/types/icon.ts
  - Define structure for icon name, display name, category, keywords
  - Support default color configuration per icon

- Implement smart icon inference in src/config/iconInference.ts
  - Create iconMappings for 25+ AI providers and cloud platforms
  - Include Claude, DeepSeek, Qwen, Kimi, Google, AWS, Azure, etc.
  - inferIconForPreset(): match provider name to icon config
  - addIconsToPresets(): batch apply icons to preset arrays
  - Support fuzzy matching for flexible name recognition

This foundation enables automatic icon assignment when users add
providers, improving visual identification in the provider list.

* feat(ui): add icon picker, color picker and provider icon components

Implement comprehensive icon selection system for provider customization:

## New Components

### ProviderIcon (src/components/ProviderIcon.tsx)
- Render SVG icons by name with automatic fallback
- Display provider initials when icon not found
- Support custom sizing via size prop
- Use dangerouslySetInnerHTML for inline SVG rendering

### IconPicker (src/components/IconPicker.tsx)
- Grid-based icon selection with visual preview
- Real-time search filtering by name and keywords
- Integration with icon metadata for display names
- Responsive grid layout (6-10 columns based on screen)

### ColorPicker (src/components/ColorPicker.tsx)
- 12 preset colors for quick selection
- Native color input for custom color picking
- Hex input field for precise color entry
- Visual feedback for selected color state

## Icon Assets (src/icons/extracted/)
- 38 high-quality SVG icons for AI providers and platforms
- Includes: OpenAI, Claude, DeepSeek, Qwen, Kimi, Gemini, etc.
- Cloud platforms: AWS, Azure, Google Cloud, Cloudflare
- Auto-generated index.ts with getIcon/hasIcon helpers
- Metadata system with searchable keywords per icon

## Build Scripts
- scripts/extract-icons.js: Extract icons from simple-icons
- scripts/generate-icon-index.js: Generate TypeScript index file

* feat(provider): integrate icon system into provider UI components

Add icon customization support to provider management interface:

## Type System Updates

### Provider Interface (src/types.ts)
- Add optional `icon` field for icon name (e.g., "openai", "anthropic")
- Add optional `iconColor` field for hex color (e.g., "#00A67E")

### Form Schema (src/lib/schemas/provider.ts)
- Extend providerSchema with icon and iconColor optional fields
- Maintain backward compatibility with existing providers

## UI Components

### ProviderCard (src/components/providers/ProviderCard.tsx)
- Display ProviderIcon alongside provider name
- Add icon container with hover animation effect
- Adjust layout spacing for icon placement
- Update translate offsets for action buttons

### BasicFormFields (src/components/providers/forms/BasicFormFields.tsx)
- Add icon preview section showing current selection
- Implement fullscreen icon picker dialog
- Auto-apply default color from icon metadata on selection
- Display provider name and icon status in preview

### AddProviderDialog & EditProviderDialog
- Pass icon fields through form submission
- Preserve icon data during provider updates

This enables users to visually distinguish providers in the list
with custom icons, improving UX for multi-provider setups.

* feat(backend): add icon fields to Provider model and default mappings

Extend Rust backend to support provider icon customization:

## Provider Model (src-tauri/src/provider.rs)
- Add `icon: Option<String>` field for icon name
- Add `icon_color: Option<String>` field for hex color
- Use serde rename `iconColor` for frontend compatibility
- Apply skip_serializing_if for clean JSON output
- Update Provider::new() to initialize icon fields as None

## Provider Defaults (src-tauri/src/provider_defaults.rs) [NEW]
- Define ProviderIcon struct with name and color fields
- Create DEFAULT_PROVIDER_ICONS static HashMap with 23 providers:
  - AI providers: OpenAI, Anthropic, Claude, Google, Gemini,
    DeepSeek, Kimi, Moonshot, Zhipu, MiniMax, Baidu, Alibaba,
    Tencent, Meta, Microsoft, Cohere, Perplexity, Mistral, HuggingFace
  - Cloud platforms: AWS, Azure, Huawei, Cloudflare
- Implement infer_provider_icon() with exact and fuzzy matching
- Add unit tests for matching logic (exact, fuzzy, case-insensitive)

## Deep Link Support (src-tauri/src/deeplink.rs)
- Initialize icon fields when creating Provider from deep link import

## Module Registration (src-tauri/src/lib.rs)
- Register provider_defaults module

## Dependencies (Cargo.toml)
- Add once_cell for lazy static initialization

This backend support enables icon persistence and future features
like auto-icon inference during provider creation.

* chore(i18n): add translations for icon picker and provider icon

Add Chinese and English translations for icon customization feature:

## Icon Picker (iconPicker)
- search: "Search Icons" / "搜索图标"
- searchPlaceholder: "Enter icon name..." / "输入图标名称..."
- noResults: "No matching icons found" / "未找到匹配的图标"
- category.aiProvider: "AI Providers" / "AI 服务商"
- category.cloud: "Cloud Platforms" / "云平台"
- category.tool: "Dev Tools" / "开发工具"
- category.other: "Other" / "其他"

## Provider Icon (providerIcon)
- label: "Icon" / "图标"
- colorLabel: "Icon Color" / "图标颜色"
- selectIcon: "Select Icon" / "选择图标"
- preview: "Preview" / "预览"

These translations support the new icon picker UI components
and provider form icon selection interface.

* style(ui): refine header layout and AppSwitcher color scheme

Improve application header and component styling:

## App.tsx Header Layout
- Wrap title and settings button in flex container with gap
- Add vertical divider between title and settings icon
- Apply responsive padding (pl-1 sm:pl-2)
- Reformat JSX for better readability (prettier)
- Fix string template formatting in className

## AppSwitcher Color Update
- Change Claude tab gradient from orange/amber to teal/emerald/green
- Update shadow color to match new teal theme
- Change hover color from orange-500 to teal-500
- Align visual style with emerald/teal brand colors

## Dialog Component Cleanup
- Remove default close button (X icon) from DialogContent
- Allow parent components to control close button placement
- Remove unused lucide-react X import

## index.css Header Border
- Add top border (2px solid) to glass-header
- Apply to both light and dark mode variants
- Improve visual separation of header area

These changes enhance visual consistency and modernize the UI
appearance with a cohesive teal color scheme.

* chore(deps): add icon library and update preset configurations

Add dependencies and utility scripts for icon system:

## Dependencies (package.json)
- Add @lobehub/icons-static-svg@1.73.0
  - High-quality SVG icon library for AI providers
  - Source for extracted icons in src/icons/extracted/
- Update pnpm-lock.yaml accordingly

## Provider Preset Updates (src/config/claudeProviderPresets.ts)
- Add optional `icon` and `iconColor` fields to ProviderPreset interface
- Apply to Anthropic Official preset as example:
  - icon: "anthropic"
  - iconColor: "#D4915D"
- Future presets can include default icon configurations

## Utility Script (scripts/filter-icons.js) [NEW]
- Helper script for filtering and managing icon assets
- Supports icon discovery and validation workflow
- Complements extract-icons.js and generate-icon-index.js

This completes the icon system infrastructure, providing all
necessary tools and dependencies for icon customization.

* refactor(ui): simplify AppSwitcher styles and migrate to local SVG icons

- Replace complex gradient animations with clean, minimal tab design
- Migrate from @lobehub/icons CDN to local SVG assets for better reliability
- Fix clippy warning in error.rs (use inline format args)
- Improve code formatting in skill service and commands
- Reduce CSS complexity in AppSwitcher component (removed blur effects and gradients)
- Update BrandIcons to use imported local SVG files instead of dynamic image loading

This improves performance, reduces external dependencies, and provides a cleaner UI experience.

* style(ui): hide scrollbars across all browsers and optimize form layout

- Hide scrollbars globally with cross-browser support:
  * WebKit browsers (Chrome, Safari, Edge): ::-webkit-scrollbar { display: none }
  * Firefox: scrollbar-width: none
  * IE 10+: -ms-overflow-style: none
- Remove custom scrollbar styling (width, colors, hover states)
- Reorganize BasicFormFields layout:
  * Move icon picker to top center as a clickable preview (80x80)
  * Change name and notes fields to horizontal grid layout (md:grid-cols-2)
  * Remove icon preview section from bottom
  * Improve visual hierarchy and space efficiency

This provides a cleaner, more modern UI with hidden scrollbars while maintaining full scroll functionality.

* refactor(layout): standardize max-width to 60rem and optimize padding structure

- Unify container max-width across components:
  * Replace max-w-4xl with max-w-[60rem] in App.tsx provider list
  * Replace max-w-5xl with max-w-[60rem] in PromptPanel
  * Move padding from header element to inner container for consistent spacing
- Optimize padding hierarchy:
  * Remove px-6 from header element, add to inner header container
  * Remove px-6 from main element, keep it only in provider list container
  * Consolidate PromptPanel padding: move px-6 from nested divs to outer container
  * Remove redundant pl-1 sm:pl-2 from logo/title area
- Benefits:
  * Consistent 60rem max-width provides better readability on wide screens
  * Simplified padding structure reduces CSS complexity
  * Cleaner responsive behavior with unified spacing rules

This creates a more maintainable and visually consistent layout system.

* refactor(ui): unify layout system with 60rem max-width and consistent spacing

- Standardize container max-width across all panels:
  * Replace max-w-4xl and max-w-5xl with unified max-w-[60rem]
  * Apply to SettingsPage, UnifiedMcpPanel, SkillsPage, and FullScreenPanel
  * Ensures consistent reading width and visual balance on wide screens

- Optimize padding hierarchy and structure:
  * Move px-6 from parent elements to content containers
  * FullScreenPanel: Add max-w-[60rem] wrapper to header, content, and footer
  * Add border separators (border-t/border-b) to header and footer sections
  * Consolidate nested padding in MCP, Skills, and Prompts panels
  * Remove redundant padding layers for cleaner CSS

- Simplify component styling:
  * MCP list items: Replace card-based layout with modern group-based design
  * Remove unnecessary wrapper divs and flatten DOM structure
  * Update card hover effects with smooth transitions
  * Simplify icon selection dialog (remove description text in BasicFormFields)

- Benefits:
  * Consistent 60rem max-width provides optimal readability
  * Unified spacing rules reduce maintenance complexity
  * Cleaner component hierarchy improves performance
  * Better responsive behavior across different screen sizes
  * More cohesive visual design language throughout the app

This creates a maintainable, scalable design system foundation.

* feat(deeplink): add Claude model fields support and enhance import dialog

- Add Claude-specific model field support in deeplink import:
  * Support model (ANTHROPIC_MODEL) - general default model
  * Support haikuModel (ANTHROPIC_DEFAULT_HAIKU_MODEL)
  * Support sonnetModel (ANTHROPIC_DEFAULT_SONNET_MODEL)
  * Support opusModel (ANTHROPIC_DEFAULT_OPUS_MODEL)
  * Backend: Update DeepLinkImportRequest struct to include optional model fields
  * Frontend: Add TypeScript type definitions for new model parameters

- Enhance deeplink demo page (deplink.html):
  * Add 5 new Claude configuration examples showcasing different model setups
  * Add parameter documentation with required/optional tags
  * Include basic config (no models), single model, complete 4-model, partial models, and third-party provider examples
  * Improve visual design with param-list component and color-coded badges
  * Add detailed descriptions for each configuration scenario

- Redesign DeepLinkImportDialog layout:
  * Switch from 3-column to compact 2-column grid layout
  * Reduce dialog width from 500px to 650px for better content display
  * Add dedicated section for Claude model configurations with blue highlight box
  * Use uppercase labels and smaller text for more information density
  * Add truncation and tooltips for long URLs
  * Improve visual hierarchy with spacing and grouping
  * Increase z-index to 9999 to ensure dialog appears on top

- Minor UI refinements:
  * Update App.tsx layout adjustments
  * Optimize McpFormModal styling
  * Refine ProviderCard and BasicFormFields components

This enables users to import Claude providers with precise model configurations via deeplink.

* feat(deeplink): add config file support for deeplink import

Support importing provider configuration from embedded or remote config files.
- Add base64 dependency for config content encoding
- Support config, configFormat, and configUrl parameters
- Make homepage/endpoint/apiKey optional when config is provided
- Add config parsing and merging logic

* feat(deeplink): enhance dialog with config file preview

Add config file parsing and preview in deep link import dialog.
- Support Base64 encoded config display
- Add config file source indicator (embedded/remote)
- Parse and display config fields by app type (Claude/Codex/Gemini)
- Mask sensitive values in config preview
- Improve dialog layout and content organization

* refactor(ui): unify dialog styles and improve layout consistency

Standardize dialog and panel components across the application.
- Update dialog background to use semantic color tokens
- Adjust FullScreenPanel max-width to 56rem for better alignment
- Add drag region and prevent body scroll in full-screen panels
- Optimize button sizes and spacing in panel headers
- Apply consistent styling to all dialog-based components

* i18n: add deeplink config preview translations

Add missing translation keys for config file preview feature.
- Add configSource, configEmbedded, configRemote labels
- Add configDetails and configUrl display strings
- Support both Chinese and English versions

* feat(deeplink): enhance test page with v3.8 config file examples

Improve deeplink test page with comprehensive config file import examples.
- Add version badge for v3.8 features
- Add copy-to-clipboard functionality for all deep links
- Add Claude config file import examples (embedded/remote)
- Add Codex config file import examples (auth.json + config.toml)
- Add Gemini config file import examples (.env format)
- Add config generator tool for easy testing
- Update UI with better styling and layout

* feat(settings): add autoSaveSettings for lightweight auto-save

Add optimized auto-save function for General tab settings.
- Add autoSaveSettings method for non-destructive auto-save
- Only trigger system APIs when values actually change
- Avoid unnecessary auto-launch and plugin config updates
- Update tests to cover new functionality

* refactor(settings): simplify settings page layout and auto-save

Reorganize settings page structure and integrate autoSaveSettings.
- Move save button inline within Advanced tab content
- Remove sticky footer for cleaner layout
- Use autoSaveSettings for General tab settings
- Simplify dialog close behavior
- Refactor ImportExportSection layout

* style(providers): optimize card layout and action button sizes

Improve provider card visual density and action buttons.
- Reduce icon button sizes for compact layout
- Adjust drag handle and icon sizes
- Tighten spacing between action buttons
- Update hover translate values for better alignment

* refactor(mcp): improve form modal layout with adaptive height editor

Restructure MCP form modal for better space utilization.
- Split form into upper form fields and lower JSON editor sections
- Add full-height mode support for JsonEditor component
- Use flex layout for editor to fill available space
- Update PromptFormPanel to use full-height editor
- Fix locale text formatting

* style: unify list item styles with semantic colors

Apply consistent styling to list items across components.
- Replace hardcoded colors with semantic tokens in MCP and Prompt list items
- Add glass effect container to EndpointSpeedTest panel
- Format code for better readability

* style: format template literals for better readability

Improve code formatting for conditional className expressions.
- Break long template literals across multiple lines
- Maintain consistent formatting in MCP form and endpoint test components

* feat(deeplink): add config merge command for preview

Expose config merging functionality to frontend for preview.
- Add merge_deeplink_config Tauri command
- Make parse_and_merge_config public for reuse
- Enable frontend to display complete config before import

* feat(deeplink): merge and display config in import dialog

Enhance import dialog to fetch and display complete config.
- Call mergeDeeplinkConfig API when config is present
- Add UTF-8 base64 decoding support for config content
- Add scrollable content area with custom scrollbar styling
- Show complete configuration before user confirms import

* i18n: add config merge error message

Add translation for config file merge error handling.

* style(deeplink): format test page HTML for better readability

Improve HTML formatting in deeplink test page.
- Format multiline attributes for better readability
- Add consistent indentation to nested elements
- Break long lines in buttons and links

* refactor(usage): improve footer layout with two-row design

Reorganize usage footer for better readability and space efficiency.
- Split into two rows: update time + refresh button (row 1), usage stats (row 2)
- Move refresh button to top row next to update time
- Remove card background for cleaner look
- Add fallback text when never updated
- Improve spacing and alignment
- Format template literals for consistency
2025-11-22 19:18:35 +08:00
Bill ZHANG 824bf796a8 fix(gemini): correct MCP server config format for Gemini CLI (#275)
Transform MCP server configurations to match Gemini CLI's expected format:
- Remove 'type' field (Gemini infers transport from field presence)
- Use 'httpUrl' field for HTTP streaming servers (type: "http")
- Keep 'url' field for SSE servers (type: "sse")
- Keep 'command' field for stdio servers unchanged

This fixes MCP servers not being recognized by Gemini CLI when
configured through cc-switch.
2025-11-22 19:18:05 +08:00
Jason b64bb6cfa1 chore: bump version to v3.7.1
Prepare for v3.7.1 maintenance release.

**Version Updates**:
- package.json: 3.7.0 → 3.7.1
- src-tauri/Cargo.toml: 3.7.0 → 3.7.1
- src-tauri/tauri.conf.json: 3.7.0 → 3.7.1
- README.md: version badge updated
- README_ZH.md: version badge updated

**CHANGELOG.md**:
- Added v3.7.1 release notes (2025-11-22)
- 3 bug fixes (Skills installation, Gemini persistence, dialog overlay)
- 2 new features (Gemini config directory, ArchLinux support)
- 3 improvements (error i18n, download timeout, code formatting)
- 1 reverted feature (auto-launch)

**Code Formatting**:
- Applied prettier to SkillsPage.tsx and skillErrorParser.ts

**Pre-Release Checks**:
 TypeScript type check passed
 Prettier format check passed
 All version numbers synchronized
2025-11-22 16:57:09 +08:00
Jason d65513ae7d docs: add v3.7.1 release documentation
Add release notes for v3.7.1 maintenance release, with v3.7.1 updates at top and complete v3.7.0 documentation below.

**v3.7.1 Updates**:
- Bug Fixes: Skills installation, Gemini config persistence, dialog overlay protection
- New Features: Gemini config directory support, ArchLinux installation
- Improvements: Skills error i18n enhancement (28+ keys), code formatting

**v3.7.0 Complete Documentation**:
- Six major features (Gemini CLI, MCP v3.7.0, Skills, Prompts, Deep Link, Conflict Detection)
- Technical statistics (152 files, +18,104 / -3,732 lines)
- Strategic positioning shift: Provider Switcher → AI CLI Management Platform

Files:
- docs/release-note-v3.7.1-zh.md (Chinese)
- docs/release-note-v3.7.1-en.md (English)
2025-11-22 16:53:00 +08:00
Jason 29e32f73f3 docs(readme): update for v3.7.0 release with six major features
Update both English and Chinese README files to reflect the major platform positioning shift and new capabilities:

**Platform Positioning**
- Update tagline: "From Provider Switcher to All-in-One AI CLI Management Platform"
- Emphasize unified management of configurations, MCP, Skills, and Prompts

**v3.7.0 Highlights**
- Gemini CLI Integration: Third supported AI CLI with dual-file config
- Claude Skills Management: GitHub integration, lifecycle management
- Prompts Management: Multi-preset system with Markdown editor
- MCP v3.7.0 Unified Architecture: SSE transport, smart parser, cross-app sync
- Deep Link Protocol: ccswitch:// for one-click config sharing
- Environment Conflict Detection: Auto-detect and resolve config conflicts

**Documentation Updates**
- Add v3.7.0 release notes link
- Expand Quick Start with Skills and Prompts management guides
- Update configuration files section with Gemini environment variables
- Add ShanDianShuo sponsor logos and information

**Technical Details**
- 18,000+ lines of new code across 6 core features
- Cross-app support for Claude Code, Codex, and Gemini CLI
- Complete usage instructions for new management systems
2025-11-22 16:05:09 +08:00
Jason eb46ac8592 Revert "feat(settings): add auto-launch on system startup feature"
This reverts commit ba336fc416.

Reason: Found issues that need to be addressed before reintroducing
the auto-launch feature. Will reimplement with fixes.
2025-11-21 23:30:56 +08:00
Jason ba336fc416 feat(settings): add auto-launch on system startup feature
Implement auto-launch functionality with proper state synchronization
and error handling across Windows, macOS, and Linux platforms.

Key changes:
- Add auto_launch module using auto-launch crate 0.5
- Define typed errors (AutoLaunchPathError, AutoLaunchEnableError, etc.)
- Sync system state with settings.json on app startup
- Only call system API when auto-launch state actually changes
- Add UI toggle in Window Settings panel
- Add i18n support for auto-launch settings (en/zh)

Implementation details:
- Settings file (settings.json) is the single source of truth
- On startup, system state is synced to match settings.json
- Error handling uses Rust type system with proper error propagation
- Frontend optimized to avoid unnecessary system API calls

Platform support:
- Windows: HKEY_CURRENT_USER registry modification
- macOS: AppleScript-based launch item (configurable to Launch Agent)
- Linux: XDG autostart desktop file
2025-11-21 23:23:35 +08:00
Jason 7fa0a7b166 feat(skills): enhance error messages with i18n support
- Add structured error format with error codes and context
- Create skillErrorParser to format errors for user-friendly display
- Add comprehensive i18n keys for all skill-related errors (zh/en)
- Extend download timeout from 15s to 60s to reduce false positives
- Fix: Pass correct error title based on operation context (load/install/uninstall)

Error improvements:
- SKILL_NOT_FOUND: Show skill directory name
- DOWNLOAD_TIMEOUT: Display repo info and timeout duration with network suggestion
- DOWNLOAD_FAILED: Show HTTP status code with specific suggestions (403/404/429)
- SKILL_DIR_NOT_FOUND: Show full path with URL check suggestion
- EMPTY_ARCHIVE: Suggest checking repository URL

Users will now see detailed error messages instead of generic "Error".
2025-11-21 16:20:01 +08:00
YoVinchen 5e54656d45 fix(skills): resolve third-party skills installation failure (#268)
- Add skills_path field to Skill struct
- Use skills_path to construct correct source path during installation
- Fix installation for repos with custom skill subdirectories
2025-11-21 15:02:01 +08:00
Jason 74969ae968 fix(dialog): prevent dialogs from closing on overlay click
Add onInteractOutside handler to DialogContent to prevent accidental
dialog closure when users click on the overlay/backdrop. This prevents
data loss in forms and improves user experience across all 11 dialog
components in the application.

Users can still close dialogs using:
- Close button (X) in the top-right corner
- Cancel/Close buttons within the dialog
- ESC key
2025-11-20 23:29:57 +08:00
Jason 1f3627add3 fix(gemini): persist settings json edits 2025-11-20 20:23:22 +08:00
YoVinchen 14ee122b27 feat(settings): add Gemini configuration directory support (#255)
* style: apply code formatting across backend and frontend

Apply cargo fmt and prettier formatting to improve code readability.
No functional changes.

Changes:
- Rust: multi-line assertion formatting (gemini_config, env_checker)
- Rust: simplify chained method calls (provider)
- TypeScript: add trailing commas to function parameters (codexProviderPresets)

* feat(settings): add Gemini configuration directory support

Add custom configuration directory support for Gemini:
- Add geminiConfigDir field to Settings type
- Extend DirectorySettings component with Gemini input
- Update useDirectorySettings hook for Gemini directory management
- Add i18n translations for Gemini directory settings
2025-11-19 21:14:43 +08:00
wugeer 7aecba14fe docs: Add installation instructions for Arch Linux (#259) 2025-11-19 19:12:00 +08:00
Jason 99b5f881e8 docs: add v3.7.0 release documentation
- Update CHANGELOG.md with v3.7.0 entry covering six major features
- Add English release notes (docs/release-note-v3.7.0-en.md)
- Add Chinese release notes (docs/release-note-v3.7.0-zh.md)

Major features documented:
- Gemini CLI integration (third app support)
- MCP v3.7.0 unified architecture
- Claude Skills management system (~2,000 lines)
- Prompts management system (~1,300 lines)
- Deep link protocol (ccswitch://)
- Environment variable conflict detection
2025-11-19 12:39:45 +08:00
Jason 286bafbd67 test: simplify boolean assertions in import_export_sync tests
Replace verbose assert_eq!(value, true) with idiomatic assert!(value) for improved readability and adherence to Rust best practices
2025-11-19 11:45:07 +08:00
Jason 6046cf8767 update screen shots 2025-11-19 11:29:19 +08:00
YoVinchen c88afa365f style: apply code formatting across backend and frontend (#252)
Apply cargo fmt and prettier formatting to improve code readability.
No functional changes.

Changes:
- Rust: multi-line assertion formatting (gemini_config, env_checker)
- Rust: simplify chained method calls (provider)
- TypeScript: add trailing commas to function parameters (codexProviderPresets)
2025-11-19 11:26:31 +08:00
farion1231 93fa5fe29a chore(release): bump version to v3.7.0 and rebrand to include Gemini CLI
- Update version from 3.6.2 to 3.7.0 across all config files
- Update project description to "All-in-One Assistant for Claude Code, Codex & Gemini CLI"
- Update Chinese description to "Claude Code / Codex / Gemini CLI 全方位辅助工具"
- Sync version in package.json, Cargo.toml, tauri.conf.json
- Update branding in README.md, README_ZH.md and i18n locale files
2025-11-19 11:20:55 +08:00
farion1231 3d31ad64af feat: add DouBaoSeed provider and remove AnyRouter preset
- Add DouBaoSeed code preview provider preset with Volcengine ARK API configuration
- Remove AnyRouter provider from both Claude and Codex preset configurations
- Clean up trailing commas in codex provider presets

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-19 11:04:44 +08:00
farion1231 bb0951552d feat: update Gemini default model and remove Google Official preset model
Updated default model from gemini-2.5-pro to gemini-3-pro-preview across:
- Provider presets (PackyCode, Custom)
- Form field placeholders
- Default configurations
- Test cases

Google Official preset now has empty env config, allowing users to choose
their own model or use application defaults, which is more appropriate for
OAuth-based authentication.

Changes:
- geminiProviderPresets.ts: updated model to gemini-3-pro-preview, removed model from Google Official
- GeminiFormFields.tsx: updated placeholder to gemini-3-pro-preview
- GeminiConfigSections.tsx: updated placeholder to gemini-3-pro-preview
- ProviderForm.tsx: updated default config to gemini-3-pro-preview
- gemini_config.rs: updated test examples to gemini-3-pro-preview
2025-11-19 10:53:33 +08:00
farion1231 00e3e6fa70 fix: read both .env and settings.json for Gemini live config
Previously, when editing a Gemini provider, only the .env file was read,
missing the settings.json file that contains MCP configuration and other
settings. This caused the config field in the edit form to be empty or
show outdated data.

This fix ensures both files are read and merged into the complete
structure { "env": {...}, "config": {...} }, matching the behavior
of Codex provider.

Fixed in:
- read_live_settings(): now reads both .env and settings.json
- import_default_config(): now reads both files when importing
2025-11-19 10:45:45 +08:00
farion1231 1ce007622e fix: resolve winreg API compatibility issue on Windows
- Update RegValue.to_string() usage to match current winreg API
  which returns String directly instead of Result
- Add conditional compilation for std::fs import (Unix only)
2025-11-19 10:06:52 +08:00
Jason 436f0e8e42 fix: sync Gemini form fields with env editor
The Gemini API key, base URL, and model inputs were not syncing with
the env editor below due to data source mismatch. The form was using
generic hooks (useApiKeyState, useBaseUrlState) that only updated
settingsConfig, while the env editor relied on geminiEnv from
useGeminiConfigState.

Changes:
- Use geminiApiKey/geminiBaseUrl from useGeminiConfigState instead of
  generic hooks
- Wrap handlers to maintain bidirectional sync between geminiEnv and
  settingsConfig
- Remove unused handleGeminiBaseUrlChange from useBaseUrlState to
  avoid naming conflicts

Now all Gemini form fields properly sync with the env editor in both
directions.
2025-11-19 09:28:53 +08:00
YoVinchen 3d69da5b66 feat: add model configuration support and fix Gemini deeplink bug (#251)
* feat(providers): add notes field for provider management

- Add notes field to Provider model (backend and frontend)
- Display notes with higher priority than URL in provider card
- Style notes as non-clickable text to differentiate from URLs
- Add notes input field in provider form
- Add i18n support (zh/en) for notes field

* chore: format code and clean up unused props

- Run cargo fmt on Rust backend code
- Format TypeScript imports and code style
- Remove unused appId prop from ProviderPresetSelector
- Clean up unused variables in tests
- Integrate notes field handling in provider dialogs

* feat(deeplink): implement ccswitch:// protocol for provider import

Add deep link support to enable one-click provider configuration import via ccswitch:// URLs.

Backend:
- Implement URL parsing and validation (src-tauri/src/deeplink.rs)
- Add Tauri commands for parse and import (src-tauri/src/commands/deeplink.rs)
- Register ccswitch:// protocol in macOS Info.plist
- Add comprehensive unit tests (src-tauri/tests/deeplink_import.rs)

Frontend:
- Create confirmation dialog with security review UI (src/components/DeepLinkImportDialog.tsx)
- Add API wrapper (src/lib/api/deeplink.ts)
- Integrate event listeners in App.tsx

Configuration:
- Update Tauri config for deep link handling
- Add i18n support for Chinese and English
- Include test page for deep link validation (deeplink-test.html)

Files: 15 changed, 1312 insertions(+)

* chore(deeplink): integrate deep link handling into app lifecycle

Wire up deep link infrastructure with app initialization and event handling.

Backend Integration:
- Register deep link module and commands in mod.rs
- Add URL handling in app setup (src-tauri/src/lib.rs:handle_deeplink_url)
- Handle deep links from single instance callback (Windows/Linux CLI)
- Handle deep links from macOS system events
- Add tauri-plugin-deep-link dependency (Cargo.toml)

Frontend Integration:
- Listen for deeplink-import/deeplink-error events in App.tsx
- Update DeepLinkImportDialog component imports

Configuration:
- Enable deep link plugin in tauri.conf.json
- Update Cargo.lock for new dependencies

Localization:
- Add Chinese translations for deep link UI (zh.json)
- Add English translations for deep link UI (en.json)

Files: 9 changed, 359 insertions(+), 18 deletions(-)

* refactor(deeplink): enhance Codex provider template generation

Align deep link import with UI preset generation logic by:
- Adding complete config.toml template matching frontend defaults
- Generating safe provider name from sanitized input
- Including model_provider, reasoning_effort, and wire_api settings
- Removing minimal template that only contained base_url
- Cleaning up deprecated test file deeplink-test.html

* style: fix clippy uninlined_format_args warnings

Apply clippy --fix to use inline format arguments in:
- src/mcp.rs (8 fixes)
- src/services/env_manager.rs (10 fixes)

* style: apply code formatting and cleanup

- Format TypeScript files with Prettier (App.tsx, EnvWarningBanner.tsx, formatters.ts)
- Organize Rust imports and module order alphabetically
- Add newline at end of JSON files (en.json, zh.json)
- Update Cargo.lock for dependency changes

* feat: add model name configuration support for Codex and fix Gemini model handling

- Add visual model name input field for Codex providers
  - Add model name extraction and update utilities in providerConfigUtils
  - Implement model name state management in useCodexConfigState hook
  - Add conditional model field rendering in CodexFormFields (non-official only)
  - Integrate model name sync with TOML config in ProviderForm

- Fix Gemini deeplink model injection bug
  - Correct environment variable name from GOOGLE_GEMINI_MODEL to GEMINI_MODEL
  - Add test cases for Gemini model injection (with/without model)
  - All tests passing (9/9)

- Fix Gemini model field binding in edit mode
  - Add geminiModel state to useGeminiConfigState hook
  - Extract model value during initialization and reset
  - Sync model field with geminiEnv state to prevent data loss on submit
  - Fix missing model value display when editing Gemini providers

Changes:
  - 6 files changed, 245 insertions(+), 13 deletions(-)
2025-11-19 09:03:18 +08:00
Jason 0ae9ed5a17 fix: resolve JSON syntax error in i18n locale files
Fixed missing closing braces in the error object of both en.json and zh.json locale files that caused Vite parse errors. The envManager.error object was not properly closed, causing the subsequent skills object to be parsed incorrectly.
2025-11-19 08:52:26 +08:00
冰子 5ff689af82 Add Gemini environment variable detection (#250)
* feat(env): add environment variable conflict detection and management

实现了系统环境变量冲突检测与管理功能:

核心功能:
- 自动检测会影响 Claude/Codex 的系统环境变量
- 支持 Windows 注册表和 Unix shell 配置文件检测
- 提供可视化的环境变量冲突警告横幅
- 支持批量选择和删除环境变量
- 删除前自动备份,支持后续恢复

技术实现:
- Rust 后端: 跨平台环境变量检测与管理
- React 前端: EnvWarningBanner 组件交互界面
- 国际化支持: 中英文界面
- 类型安全: 完整的 TypeScript 类型定义

* refactor(env): remove unused imports and function

Remove unused HashMap and PathBuf imports, and delete the unused get_source_description function to clean up the code.

* feat: Add Gemini environment variable detection
2025-11-19 08:33:02 +08:00
冰子 b9412ece0b 添加Claude和Codex环境变量检查 (#242)
* feat(env): add environment variable conflict detection and management

实现了系统环境变量冲突检测与管理功能:

核心功能:
- 自动检测会影响 Claude/Codex 的系统环境变量
- 支持 Windows 注册表和 Unix shell 配置文件检测
- 提供可视化的环境变量冲突警告横幅
- 支持批量选择和删除环境变量
- 删除前自动备份,支持后续恢复

技术实现:
- Rust 后端: 跨平台环境变量检测与管理
- React 前端: EnvWarningBanner 组件交互界面
- 国际化支持: 中英文界面
- 类型安全: 完整的 TypeScript 类型定义

* refactor(env): remove unused imports and function

Remove unused HashMap and PathBuf imports, and delete the unused get_source_description function to clean up the code.
2025-11-18 23:44:44 +08:00
YoVinchen ec303544ca Feat/claude skills management (#237)
* feat(skills): add Claude Skills management feature

Implement complete Skills management system with repository discovery,
installation, and lifecycle management capabilities.

Backend:
- Add SkillService with GitHub integration and installation logic
- Implement skill commands (list, install, uninstall, check updates)
- Support multiple skill repositories with caching

Frontend:
- Add Skills management page with repository browser
- Create SkillCard and RepoManager components
- Add badge, card, table UI components
- Integrate Skills API with Tauri commands

Files: 10 files changed, 1488 insertions(+)

* feat(skills): integrate Skills feature into application

Integrate Skills management feature with complete dependency updates,
configuration structure extensions, and internationalization support.

Dependencies:
- Add @radix-ui/react-visually-hidden for accessibility
- Add anyhow, zip, serde_yaml, tempfile for Skills backend
- Enable chrono serde feature for timestamp serialization

Backend Integration:
- Extend MultiAppConfig with SkillStore field
- Implement skills.json migration from legacy location
- Register SkillService and skill commands in main app
- Export skill module in commands and services

Frontend Integration:
- Add Skills page route and dialog in App
- Integrate Skills UI with main navigation

Internationalization:
- Add complete Chinese translations for Skills UI
- Add complete English translations for Skills UI

Code Quality:
- Remove redundant blank lines in gemini_mcp.rs
- Format log statements in mcp.rs

Tests:
- Update import_export_sync tests for SkillStore
- Update mcp_commands tests for new structure

Files: 16 files changed, 540 insertions(+), 39 deletions(-)

* style(skills): improve SkillsPage typography and spacing

Optimize visual hierarchy and readability of Skills page:
- Reduce title size from 2xl to lg with tighter tracking
- Improve description spacing and color contrast
- Enhance empty state with better text hierarchy
- Use explicit gray colors for better dark mode support

* feat(skills): support custom subdirectory path for skill scanning

Add optional skillsPath field to SkillRepo to enable scanning skills
from subdirectories (e.g., "skills/") instead of repository root.

Changes:
- Backend: Add skillsPath field with subdirectory scanning logic
- Frontend: Add skillsPath input field and display in repo list
- Presets: Add cexll/myclaude repo with skills/ subdirectory
- Code quality: Fix clippy warnings (dedup logic, string formatting)

Backward compatible: skillsPath is optional, defaults to root scanning.

* refactor(skills): improve repo manager dialog layout

Optimize dialog structure with fixed header and scrollable content:
- Add flexbox layout with fixed header and scrollable body
- Remove outer border wrapper for cleaner appearance
- Match SkillsPage design pattern for consistency
- Improve UX with better content hierarchy
2025-11-18 22:05:54 +08:00
Jason 023726c59d docs: add Codex MCP raw TOML refactor plan
Add comprehensive implementation plan for v3.7.0 refactor that
separates Codex MCP configuration from unified MCP structure.

Key design decisions:
- Codex MCP stored as raw TOML string in codexMcp.rawToml
- Unified MCP (mcp.servers) only supports Claude/Gemini
- Complete isolation: no apps.codex in unified structure
- Migration clears mcp.codex.servers to prevent pollution

Architecture improvements:
- Single responsibility: each data source has one purpose
- No priority conflicts: completely independent data paths
- Simplified switching logic: no conditional branches
- UI constraints: Tab1 limited to claude/gemini only

Implementation phases:
- Phase 0: Setup (0.5d)
- Phase 1: Backend foundation (1.5d)
- Phase 2: Command layer (1d)
- Phase 3: Switching logic (1d)
- Phase 4: Frontend API (0.5d)
- Phase 5: UI implementation (2d)
- Phase 6: Enhancements (1d, optional)
- Phase 7: Testing & docs (1d)

Total: 8.5 days (MVP: 7 days)

Addresses TOML-JSON conversion data loss issue by preserving
raw TOML format for Codex while maintaining structured approach
for Claude/Gemini.
2025-11-18 16:08:31 +08:00
Jason 8d2c067814 feat(window): center window on screen by default
Add `center: true` to main window configuration to improve initial
window positioning on Windows and other platforms. This addresses
user feedback about the window appearing in the top-left corner.
2025-11-17 23:25:13 +08:00
Jason a00eb764f7 fix(provider): sync MCP config for all apps on provider switch
Previously, only Codex provider switches triggered MCP synchronization,
which could cause MCP configuration loss when switching Claude or Gemini
providers.

Changes:
- Enable MCP sync for all app types (Claude, Codex, Gemini) during provider switch
- Migrate to v3.7.0 unified MCP sync mechanism using McpService::sync_all_enabled()
- Replace app-specific sync_enabled_to_codex() with unified sync for all apps
- Remove unused mcp module import

This ensures MCP servers remain properly configured across all applications
after provider switches, preventing configuration loss.
2025-11-17 23:17:21 +08:00
Jason 67bd8f5c11 fix(mcp): correct Codex MCP configuration format to [mcp_servers]
BREAKING CHANGE: The [mcp.servers] format was completely incorrect and not
any official Codex format. The only correct format is [mcp_servers] at the
top level of config.toml.

Changes:
- Remove incorrect [mcp.servers] nested table support
- Always use [mcp_servers] top-level table (official Codex format)
- Auto-migrate and cleanup erroneous [mcp.servers] entries on write
- Preserve error-tolerant import for migrating old incorrect configs
- Simplify sync logic by removing format selection branches (~60 lines)
- Update all documentation and tests to reflect correct format
- Add warning logs when detecting and cleaning incorrect format

Backend (Rust):
- mcp.rs: Simplify sync_enabled_to_codex by removing Target enum
- mcp.rs: sync_single_server_to_codex now always uses [mcp_servers]
- mcp.rs: remove_server_from_codex cleans both locations
- mcp.rs: Update import_from_codex comments to clarify format status
- tests: Rename test to sync_enabled_to_codex_migrates_erroneous_*
- tests: Update assertions to verify migration behavior

Frontend (TypeScript):
- tomlUtils.ts: Prioritize [mcp_servers] format in parsing
- tomlUtils.ts: Update error messages to guide correct format

Documentation:
- README.md: Correct MCP format reference to [mcp_servers]
- CLAUDE.md: Add comprehensive format specification with examples

All 79 tests pass. This ensures backward compatibility while enforcing
the correct Codex official standard going forward.

Refs: https://github.com/openai/codex/issues/3441
2025-11-17 22:57:04 +08:00
Jason 3051743bd3 feat(mcp): support extended fields for Codex TOML conversion
## Problem
Previously, the Codex MCP configuration used a whitelist-only approach
for TOML conversion, which caused custom fields (like `timeout`,
`startup_timeout_ms`, etc.) to be silently dropped during JSON → TOML
conversion. Only Claude and Gemini (JSON format) could preserve
arbitrary fields.

## Solution
Implemented a three-tier field handling strategy:

1. **Core fields** (type, command, args, url, headers, env, cwd)
   - Strong-typed manual processing (existing behavior preserved)

2. **Extended fields** (19 common optional fields)
   - White-listed fields with automatic type conversion:
     - General: timeout, timeout_ms, startup_timeout_ms/sec,
       connection_timeout, read_timeout, debug, log_level, disabled
     - stdio: shell, encoding, working_dir, restart_on_exit,
       max_restart_count
     - http/sse: retry_count, max_retry_attempts, retry_delay,
       cache_tools_list, verify_ssl, insecure, proxy

3. **Custom fields** (generic converter)
   - Automatic type inference for:
     - String → TOML String
     - Number (i64/f64) → TOML Integer/Float
     - Boolean → TOML Boolean
     - Simple arrays → TOML Array
     - Shallow objects (string values only) → TOML Inline Table
   - Unsupported types (null, mixed arrays, nested objects) are
     gracefully skipped with debug logging

## Changes

### Core Implementation
- **json_value_to_toml_item()** (mcp.rs:843-947)
  Generic JSON → TOML value converter with smart type inference

- **json_server_to_toml_table()** (mcp.rs:949-1072)
  Refactored to use three-tier strategy, processes all fields beyond
  core whitelist

- **build_servers_table()** (mcp.rs:616-633)
  Now reuses the generic converter for consistency

- **import_from_codex()** (mcp.rs:432-605)
  Extended with generic TOML → JSON converter for bidirectional
  field preservation

### Logging
- DEBUG: Extended field conversions
- INFO: Custom field conversions
- WARN: Skipped unsupported types (with reason)

## Testing
-  All 24 integration tests pass
-  Compilation clean (zero errors)
-  Backward compatible with existing configs

## Design Decision: No Auto Field Mapping
Explicitly NOT implementing automatic field mapping (e.g., Claude's
`timeout` → Codex's `startup_timeout_ms`) due to:
- Unit ambiguity (seconds vs milliseconds)
- Semantic ambiguity (same field name, different meanings)
- Risk of data corruption (30s → 30ms causes immediate timeout)
- Breaks user expectation of "what you see is what you get"

Recommendation: Use application-specific field names as documented.

## Example
User adds `timeout: 30` in MCP panel:

**Claude/Gemini** (~/.claude.json):
```json
{"mcpServers": {"srv": {"timeout": 30}}}
```

**Codex** (~/.codex/config.toml):
```toml
[mcp.servers.srv]
timeout = 30  #  Now preserved!
```

Fixes the whitelist limitation reported in user feedback.
2025-11-17 17:23:09 +08:00
qixing-jk 883cf0346b fix(password input): disable Edge/IE reveal and clear buttons (#232)
Hide default password reveal and clear controls in Microsoft browsers
for improved consistency and security.
2025-11-17 15:30:46 +08:00
Jason 1805ed586e fix(mcp): improve format/submit UX and fix validation errors
Fixes two critical issues with the MCP JSON input:
1. Format button failed with wrapped format like "server": {...}
2. Submit button failed despite input validation passing

Changes:
- Updated formatJSON to use smart parser (supports wrapped format)
- Simplified submit validation logic (removed redundant validateJsonConfig call)
- Improved UX: input preserves original format, cleanup happens on format/submit
  * Prevents confusing "instant disappearance" when pasting wrapped JSON
  * Auto-fills ID/Name fields while keeping input unchanged
  * Format button now strips wrapper key and formats cleanly
  * Submit button correctly extracts config regardless of format

Code quality:
- Reduced code by 5 lines (20 changes, 11 insertions, 16 deletions)
- Consistent use of parseSmartMcpJson across all JSON operations
- No type errors introduced
2025-11-16 20:40:16 +08:00
Jason 98a1305684 feat(mcp): add smart JSON parser for flexible input formats
Support multiple MCP configuration input formats:
- Pure config object: { "command": "npx", ... }
- Key-value pair fragment: "server-name": { "command": "npx", ... }
- Wrapped object: { "server-name": { "command": "npx", ... } }

The parser automatically:
- Detects and wraps JSON fragments into complete objects
- Extracts server name from single-key objects
- Auto-fills ID and Name fields when applicable
- Formats the config for display

This improves UX by allowing users to paste configs directly from
.claude.json or .codex/config.toml without manual editing.
2025-11-16 20:08:04 +08:00
Jason f79efb86cd refactor(mcp): improve form label layout and simplify text
- Simplify config label from "Full JSON configuration or use" to "Full JSON Configuration"
- Align wizard button to the right using justify-between layout
- Apply same pattern to TOML configuration label
- Improve visual balance with cleaner left-right alignment
2025-11-16 16:59:23 +08:00
Jason ed59420a83 feat(mcp): enhance form UX with default apps and JSON formatter
- Enable all apps (Claude, Codex, Gemini) by default when adding MCP servers
- Improve config label with clearer wording: "Full JSON configuration or use [Config Wizard]"
- Add JSON format button to beautify configuration with 2-space indentation
- Update tests to reflect new default behavior
- Clean up redundant explicit prop passing

This provides a more streamlined experience by enabling all apps out of the box
and making it easier to format JSON configurations.
2025-11-16 16:50:07 +08:00
Jason bfc27349b3 feat(mcp): add SSE (Server-Sent Events) transport type support
Add comprehensive support for SSE transport type to MCP server configuration,
enabling real-time streaming connections alongside existing stdio and http types.

Backend Changes:
- Add SSE type validation in mcp.rs validate_server_spec()
- Extend Codex TOML import/export to handle SSE servers
- Update claude_mcp.rs legacy API for backward compatibility
- Unify http/sse handling in json_server_to_toml_table()

Frontend Changes:
- Extend McpServerSpec type definition to include "sse"
- Add SSE radio button to configuration wizard UI
- Update wizard form logic to handle SSE url and headers
- Add SSE validation in McpFormModal submission

Validation & Error Handling:
- Add SSE support in useMcpValidation hook (TOML/JSON)
- Extend tomlUtils normalizeServerConfig for SSE parsing
- Update Zod schemas (common.ts, mcp.ts) with SSE enum
- Add SSE error message mapping in errorUtils

Internationalization:
- Add "typeSse" translations (zh: "sse", en: "sse")

Tests:
- Add SSE validation test cases in useMcpValidation.test.tsx

SSE Configuration Format:
{
  "type": "sse",
  "url": "https://api.example.com/sse",
  "headers": { "Authorization": "Bearer token" }
}
2025-11-16 16:15:17 +08:00
Jason 4fc7413ffa refactor(codex): simplify custom template with minimal config
**Changes:**
- Remove all comments from custom template (align with Claude/Gemini)
- Remove base_url field from template (user fills in form instead)
- Simplify getCodexCustomTemplate() - no locale parameter needed
- Keep preset configurations unchanged with their baseUrl values

**Template now contains only:**
- model_provider, model, model_reasoning_effort
- disable_response_storage
- [model_providers.custom] section with minimal fields

**Benefits:**
-  Cleaner, more focused template
-  Consistent with other apps (no comments)
-  Forces users to fill base_url via form field
-  Reduced template size from 35+ lines to 12 lines

Net change: -74 lines (codexTemplates.ts)
2025-11-16 13:36:33 +08:00
Jason 12112e9d7d refactor(codex): extract template to config with i18n support
**Changes:**
- Create src/config/codexTemplates.ts with getCodexCustomTemplate factory
- Support both Chinese and English templates based on i18n.language
- Remove 70 lines of duplicated template strings from ProviderForm.tsx
- Update both useEffect and handlePresetChange to use template factory
- Clean up unused "Custom (Blank Template)" preset entry

**Benefits:**
-  Eliminates code duplication (35-line template repeated twice)
-  Adds internationalization support for English users
-  Follows project architecture (templates in config/ directory)
-  Improves maintainability (single source of truth)
-  Net reduction: 34 lines (81 additions, 115 deletions)

**Technical Details:**
- Template selection logic: (i18n.language || "zh").startsWith("zh") ? "zh" : "en"
- Templates are identical except for comments language
- Both auth and config are returned as a single CodexTemplate object

Addresses DRY principle violation and architectural concerns identified
in code review.
2025-11-16 13:23:53 +08:00
Jason 6a6980c82c refactor(codex): remove configuration wizard and unify provider setup experience
- Remove CodexQuickWizardModal component (~300 lines)
- Add "Custom (Blank Template)" preset with annotated TOML template
- Unify configuration experience across Claude/Codex/Gemini
- Remove wizard-related i18n keys, keep apiUrlLabel for CodexFormFields
- Simplify component integration by removing wizard state management

This change reduces code complexity by ~250 lines while providing better
user education through commented configuration templates in Chinese.

Users can now:
1. Select "Custom (Blank Template)" preset
2. See annotated TOML template with inline documentation
3. Follow step-by-step comments to configure custom providers

BREAKING CHANGE: Configuration wizard UI removed, replaced with template-based approach
2025-11-16 12:29:18 +08:00
Jason 031ea3a58f style(mcp): refine panel layout for better visual hierarchy and compactness
- Replace checkboxes with toggle switches for app selection (more intuitive for enable/disable actions)
- Change switch color from blue to emerald to match MCP button theme
- Stack app options vertically with labels on left to save horizontal space
- Reduce panel width from max-w-4xl to max-w-3xl for more compact design
- Move docs button next to server name for better information grouping
2025-11-16 11:31:27 +08:00
Jason 9d431cc7ae style(ui): migrate color scheme to macOS native design system
- Update primary blue from Linear style (#3498db) to macOS system blue (#0A84FF)
- Align gray scale with macOS dark mode palette (#1C1C1E, #2C2C2E, etc.)
- Adjust dark mode background to match macOS systemBackground (HSL 240 5% 12%)
- Refine scrollbar colors to use native macOS gray tones
- Remove transparent titleBarStyle for better stability
2025-11-16 10:44:15 +08:00
Jason 685a1138e4 refactor(mcp): complete form refactoring for unified MCP management
Complete the v3.7.0 MCP refactoring by updating the form layer to match
the unified architecture already implemented in data/service/API layers.

**Breaking Changes:**
- Remove confusing `appId` parameter from McpFormModal
- Replace with `defaultFormat` (json/toml) and `defaultEnabledApps` (array)

**Form Enhancements:**
- Add app enablement checkboxes (Claude/Codex/Gemini) directly in the form
- Smart defaults: new servers default to Claude enabled, editing preserves state
- Support "draft" mode: servers can be created without enabling any apps

**Architecture Improvements:**
- Eliminate semantic confusion: format selection separate from app targeting
- One-step workflow: configure and enable apps in single form submission
- Consistent with unified backend: `apps: { claude, codex, gemini }`

**Testing:**
- Update test mocks to use `useUpsertMcpServer` hook
- Add test case for creating servers with no apps enabled
- Fix parameter references from `appId` to `defaultFormat`

**i18n:**
- Add `mcp.form.enabledApps` translation (zh/en)
- Add `mcp.form.noAppsWarning` translation (zh/en)

This completes the MCP management refactoring, ensuring all layers
(data, service, API, UI) follow the same unified architecture pattern.
2025-11-15 23:47:35 +08:00
Jason 154ff4c819 feat(config): unify common config snippets persistence across all apps
- Add unified `common_config_snippets` structure to MultiAppConfig
- Implement `get_common_config_snippet` and `set_common_config_snippet` commands
- Replace localStorage with config.json persistence for Codex and Gemini
- Auto-migrate legacy `claude_common_config_snippet` to new unified structure
- Deprecate individual API methods in favor of unified interface
- Add automatic migration from localStorage on first load

BREAKING CHANGE: Common config snippets now stored in unified `common_config_snippets` object instead of separate fields
2025-11-15 19:52:49 +08:00
Jason 2540f6ba08 refactor(prompt): optimize auto-import logic with unified save
- Change auto_import_prompt_if_exists return type to Result<bool>
- Remove redundant save() calls by introducing updated flag
- Eliminate unnecessary clone() in app type iterations
- Remove semantic contradiction in let _ = ...? pattern
- Improve code semantics and maintainability

Performance: Reduce disk I/O by 50%+ through unified save logic
2025-11-15 16:47:08 +08:00
YoVinchen d32ceb9b80 fix(prompt): correct live file backfill priority in enable flow (#225)
Fix backfill logic in prompt enable workflow:
- Prioritize backfilling live file content to currently enabled prompt (prevent data loss)
- Create backup only when no enabled prompt exists and content is new (avoid duplicate backups)
- Implement staged persistence (save after backfill + save after enable)
- Add explicit logging for backfill/backup operations

Also simplify string formatting in prompt_files.rs with inline format strings.
2025-11-15 16:09:00 +08:00
Jason e11c7d84cd test(mcp): update import tests for v3.7.0 unified structure
- Fix import_from_claude_merges_into_config: check unified mcp.servers
- Fix import_from_codex_adds_servers_from_mcp_servers_table: verify apps.codex enabled
- Fix import_from_codex_merges_into_existing_entries: test smart merge preserves existing config
- Replace cc_switch_lib::app_config:: with public exports (McpServer, McpApps)

All 24 import_export_sync tests now passing.
2025-11-14 23:39:34 +08:00
Jason ea8f2095e2 fix(mcp): migrate import functions to unified v3.7.0 structure
- Rewrite import_from_claude/codex/gemini to write directly to mcp.servers
- Implement skip-on-error strategy for fault tolerance (single invalid item no longer aborts entire batch)
- Smart merge logic: existing servers only enable corresponding app, preserve other configs
- Remove deprecated markers from service layer
- Export McpApps type for test usage
- Update mcp_commands tests to use unified structure

Fixes runtime import issue where data was written to legacy structure (mcp.claude/codex.servers)
but unified panel reads from new structure (mcp.servers), causing "imported but invisible" bug.
2025-11-14 23:33:54 +08:00
Jason 09f80d82bc fix(mcp): initialize McpRoot with v3.7.0 unified structure by default
Problem:
- On first launch, McpRoot::default() created `servers: None`
- McpService::get_all_servers() would incorrectly return error "old structure detected"
- This confused new users who had no legacy MCP data

Root Cause:
- Derived Default trait sets Option<T> fields to None
- New installations should start with v3.7.0 structure immediately

Solution:
- Explicitly implement Default for McpRoot
- Initialize `servers: Some(HashMap::new())` for v3.7.0+ structure
- Legacy fields (claude/codex/gemini) remain empty, only used for deserializing old configs

Impact:
- First-time users get correct v3.7.0 structure immediately
- migrate_mcp_to_unified() correctly detects already-migrated state
- No false "old structure" errors on fresh installs
2025-11-14 22:55:46 +08:00
Jason 2f18d6ec00 refactor(mcp): complete v3.7.0 cleanup - remove legacy code and warnings
This commit finalizes the v3.7.0 unified MCP architecture migration by
removing all deprecated code paths and eliminating compiler warnings.

Frontend Changes (~950 lines removed):
- Remove deprecated components: McpPanel, McpListItem, McpToggle
- Remove deprecated hook: useMcpActions
- Remove unused API methods: importFrom*, syncEnabledTo*, syncAllServers
- Simplify McpFormModal by removing dual-mode logic (unified/legacy)
- Remove syncOtherSide checkbox and conflict detection
- Clean up unused imports and state variables
- Delete associated test files

Backend Changes (~400 lines cleaned):
- Remove unused Tauri commands: import_mcp_from_*, sync_enabled_mcp_to_*
- Delete unused Gemini MCP functions: get_mcp_status, upsert/delete_mcp_server
- Add #[allow(deprecated)] to compatibility layer commands
- Add #[allow(dead_code)] to legacy helper functions for future migration
- Simplify boolean expression in mcp.rs per Clippy suggestion

API Deprecation:
- Mark legacy APIs with @deprecated JSDoc (getConfig, upsertServerInConfig, etc.)
- Preserve backward compatibility for v3.x, planned removal in v4.0

Verification:
-  Zero TypeScript errors (pnpm typecheck)
-  Zero Clippy warnings (cargo clippy)
-  All code formatted (prettier + cargo fmt)
-  Builds successfully

Total cleanup: ~1,350 lines of code removed/marked
Breaking changes: None (all legacy APIs still functional)
2025-11-14 22:43:25 +08:00
Jason fafca841cb refactor(frontend): remove redundant 'Sync All' button from MCP panel
All MCP operations already auto-sync to live configs:
- upsert_server() → sync_server_to_apps()
- toggle_app() → sync_server_to_app() or remove_server_from_app()
- delete_server() → remove_server_from_all_apps()

The manual 'Sync All' button was redundant and could confuse users
into thinking they need to manually sync after each change.

Changes:
- Remove 'Sync All' button from UnifiedMcpPanel header
- Remove useSyncAllMcpServers hook
- Remove handleSyncAll function and syncAllMutation state
- Remove RefreshCw icon import
- Remove sync-related i18n translations (en/zh)

Note: Backend sync_all_mcp_servers command remains for potential
future use (e.g., recovery tool), but is no longer exposed in UI.
2025-11-14 15:52:01 +08:00
Jason f4b8aed29a refactor(frontend): remove MCP import functionality for v3.7.0
Auto-migration at startup is sufficient for upgrading from v3.6.x.
Manual import adds unnecessary complexity since:
- Gemini MCP support launches with v3.7.0 (no legacy data)
- Existing Claude/Codex MCP configs are auto-migrated on first run
- All MCP management should happen within CC Switch

Changes:
- Remove McpImportDialog component
- Remove "Import" button from UnifiedMcpPanel
- Remove import-related i18n translations (en/zh)
- Simplify user experience with single management interface

Note: Backend import commands (import_mcp_from_*) remain for
backward compatibility but are no longer exposed in UI.
2025-11-14 15:47:04 +08:00
Jason 9663b4251e feat(frontend): add MCP import dialog for v3.7.0
Implement import functionality to migrate MCP servers from existing configs:

**New Component:**
- src/components/mcp/McpImportDialog.tsx: Import dialog with card-based source selection

**Features:**
- Import from Claude (~/.claude/claude.json or settings.json)
- Import from Codex (~/.codex/config.toml)
- Import from Gemini (config file)
- Card-based UI with icons and descriptions
- Loading states with spinner animation
- Auto-refresh after successful import

**Integration:**
- Add import button to UnifiedMcpPanel header
- Handle import completion with refetch
- Toast notifications for success/info/error cases

**I18n:**
- Add mcp.unifiedPanel.import namespace (zh/en)
- Import button, dialog title, descriptions
- Success/error messages with count interpolation

**UX:**
- Smart disable: other sources disabled during import
- Clear feedback: count of imported servers
- Friendly messages: "No servers found" when empty

TypeScript type check passes 
2025-11-14 15:29:16 +08:00
Jason 9e8abf5f26 feat(frontend): implement unified MCP panel for v3.7.0
Complete Phase 3 (P0) frontend implementation for unified MCP management:

**New Files:**
- src/hooks/useMcp.ts: React Query hooks for unified MCP operations
- src/components/mcp/UnifiedMcpPanel.tsx: Unified MCP management panel
- src/components/ui/checkbox.tsx: Checkbox component from shadcn/ui

**Features:**
- Unified panel with three-column layout: server info + app checkboxes + actions
- Multi-app control: Claude/Codex/Gemini checkboxes for each server
- Real-time stats: Show enabled server counts per app
- Full CRUD operations: Add, edit, delete, sync all servers

**Integration:**
- Replace old app-specific McpPanel with UnifiedMcpPanel in App.tsx
- Update McpFormModal to support unified mode with apps field
- Add i18n support: mcp.unifiedPanel namespace (zh/en)

**Type Safety:**
- Ensure McpServer.apps field always initialized
- Fix all test files to include apps field
- TypeScript type check passes 

**Architecture:**
- Single source of truth: mcp.servers manages all MCP configs
- Per-server app control: apps.claude/codex/gemini boolean flags
- Backward compatible: McpFormModal supports both unified and legacy modes

Next: P1 tasks (import dialogs, sub-components, tests)
2025-11-14 15:24:48 +08:00
Jason 32a6de074c docs: add comprehensive v3.7.0 unified MCP refactor plan
## Document Overview
Created detailed refactoring plan document covering:
- Project goals and architecture changes
- Data structure migration strategy (v3.6.x → v3.7.0)
- Complete development progress tracking
- Frontend remaining tasks roadmap
- Testing plan and delivery checklist

## Completed Work (Documented)
 Phase 1: Backend data structure migration
  - McpApps, McpServer, McpRoot definitions
  - Automatic migration logic
  - Integration into load() method

 Phase 2: Backend services refactor
  - Complete McpService rewrite
  - Single-server sync functions
  - New Tauri commands
  - Backward compatibility layer
  - All compilation errors fixed

 Phase 3 (Partial): Frontend types and API
  - Updated TypeScript types (McpApps, McpServer)
  - New API methods (getAllServers, toggleApp, etc.)

## Remaining Work (Documented)
📝 React Query hooks (useMcp.ts)
📝 UnifiedMcpPanel component
📝 i18n translations
📝 Main app integration
📝 Sub-components (table, form, import dialog)

## Key Highlights
- Detailed UI wireframes and component structure
- Step-by-step migration flow documentation
- Risk assessment and mitigation strategies
- Technical insights and architecture benefits

Related: v3.7.0 unified MCP management
2025-11-14 15:09:22 +08:00
Jason ac09551563 feat(frontend): add unified MCP types and API layer for v3.7.0
## Type Definitions
- Update McpServer interface with new apps field (McpApps)
- Add McpApps interface for multi-app enable state
- Add McpServersMap type for server collections
- Mark enabled field as deprecated (use apps instead)
- Maintain backward compatibility with optional fields

## API Layer Updates
- Add unified MCP management methods to mcpApi:
  * getAllServers() - retrieve all servers with apps state
  * upsertUnifiedServer() - add/update server with apps
  * deleteUnifiedServer() - remove server
  * toggleApp() - enable/disable server for specific app
  * syncAllServers() - sync all enabled servers to live configs
- Import new McpServersMap type

## Code Organization
- Keep all types in src/types.ts (removed duplicate types/mcp.ts)
- Follow existing project structure conventions

Related: v3.7.0 unified MCP management
2025-11-14 13:01:47 +08:00
Jason 7ae2a9f556 fix(mcp): resolve compilation errors and add backward compatibility
## Compilation Fixes
- Add deprecated compatibility methods to McpService:
  * get_servers() - filters servers by app
  * set_enabled() - delegates to toggle_app()
  * sync_enabled() - syncs enabled servers for specific app
  * import_from_claude/codex/gemini() - wraps mcp:: functions
- Fix toml_edit type conversion in sync_single_server_to_codex():
  * Add json_server_to_toml_table() helper function
  * Manually construct toml_edit::Table instead of invalid serde conversion
- Fix get_codex_config_path() calls (returns PathBuf, not Result)
- Update upsert_mcp_server_in_config() to work with unified structure:
  * Converts old per-app API to unified McpServer structure
  * Preserves existing server data when updating
  * Supports sync_other_side parameter for multi-app enable
- Update delete_mcp_server_in_config() to ignore app parameter

## Backward Compatibility
- All old v3.6.x commands continue to work with deprecation warnings
- Frontend migration can be done incrementally
- Old commands transparently use new unified backend

## Status
 Backend compiles successfully (cargo check passes)
⚠️ 16 warnings (8 deprecation + 8 unused functions - expected)
2025-11-14 12:57:14 +08:00
Jason c985db8f3d feat(mcp): implement unified MCP management for v3.7.0
BREAKING CHANGE: Migrate from app-specific MCP storage to unified structure

## Phase 1: Data Structure Migration
- Add McpApps struct with claude/codex/gemini boolean fields
- Add McpServer struct for unified server definition
- Add migration logic in MultiAppConfig::migrate_mcp_to_unified()
- Integrate automatic migration into MultiAppConfig::load()
- Support backward compatibility through Optional fields

## Phase 2: Backend Services Refactor
- Completely rewrite services/mcp.rs for unified management:
  * get_all_servers() - retrieve all MCP servers
  * upsert_server() - add/update unified server
  * delete_server() - remove server
  * toggle_app() - enable/disable server for specific app
  * sync_all_enabled() - sync to all live configs
- Add single-server sync functions to mcp.rs:
  * sync_single_server_to_claude/codex/gemini()
  * remove_server_from_claude/codex/gemini()
- Add read_mcp_servers_map() to claude_mcp.rs and gemini_mcp.rs
- Add new Tauri commands to commands/mcp.rs:
  * get_mcp_servers, upsert_mcp_server, delete_mcp_server
  * toggle_mcp_app, sync_all_mcp_servers
- Register new commands in lib.rs

## Migration Strategy
- Detects old structure (mcp.claude/codex/gemini.servers)
- Merges into unified mcp.servers with apps markers
- Handles conflicts by merging enabled apps
- Clears old structures after migration
- Saves migrated config automatically

## Known Issues
- Old commands still need compatibility layer (WIP)
- toml_edit type conversion needs fixing (WIP)
- Frontend not yet implemented (Phase 3 pending)

Related: v3.6.2 -> v3.7.0
2025-11-14 12:51:24 +08:00
Jason c7b235bb98 fix(i18n): add missing Gemini MCP panel title and fix ternary logic
- Add mcp.geminiTitle to both zh.json and en.json
- Fix McpPanel title logic to handle all three apps (claude/codex/gemini)
- Previous logic would incorrectly display codexTitle for gemini
2025-11-14 10:35:55 +08:00
Jason 1616c63c0b feat(gemini): implement full MCP management functionality
- Add gemini_mcp.rs module for Gemini MCP file I/O operations
- Implement sync_enabled_to_gemini to export enabled MCPs to ~/.gemini/settings.json
- Implement import_from_gemini to import MCPs from Gemini config
- Add Gemini sync logic in services/mcp.rs (upsert_server, delete_server, set_enabled)
- Register Tauri commands for Gemini MCP sync and import
- Update frontend API calls and McpPanel to support Gemini

Fixes the issue where adding MCP servers in Gemini tab would not sync to ~/.gemini/settings.json
2025-11-14 10:02:27 +08:00
Jason 146b42fb68 feat(gemini): add config.json editor and common config functionality
Implements dual-editor pattern for Gemini providers, following the Codex architecture:
- Environment variables (.env format) editor
- Extended configuration (config.json) editor with common config support

New Components:
- GeminiConfigSections: Separate sections for env and config editing
- GeminiCommonConfigModal: Modal for editing common config snippets

New Hooks:
- useGeminiConfigState: Manages env/config separation and conversion
  - Converts between .env string format and JSON object
  - Validates JSON config structure
  - Extracts API Key and Base URL from env
- useGeminiCommonConfig: Handles common config snippets
  - Deep merge algorithm for combining configs
  - Remove common config logic for toggling off
  - localStorage persistence for snippets

Features:
- Format buttons for both env and config editors
- Common config toggle with deep merge/remove
- Error validation and display
- Auto-open modal on common config errors

Configuration Structure:
{
  "env": {
    "GOOGLE_GEMINI_BASE_URL": "https://...",
    "GEMINI_API_KEY": "sk-...",
    "GEMINI_MODEL": "gemini-2.5-pro"
  },
  "config": {
    "timeout": 30000,
    "maxRetries": 3
  }
}

This brings Gemini providers to feature parity with Claude and Codex.
2025-11-14 08:32:30 +08:00
Jason 0ea434a485 fix(i18n): deduplicate category labels by reusing providerForm keys
- Change ProviderForm to use providerForm.category* instead of providerPreset.category*
- Remove duplicate category keys from providerPreset namespace in both zh.json and en.json
- Fix naming inconsistency: use categoryAggregation (not categoryAggregator)
- Fixes issue where English UI would show Chinese defaultValue fallbacks

This ensures single source of truth for category labels and improves maintainability.
2025-11-14 08:31:09 +08:00
Jason 21fd7cc9fd feat: migrate Claude common config snippet from localStorage to config.json
Migrate the Claude common config snippet storage from browser localStorage
to the persistent config.json file for better cross-device sync and backup support.

**Backend Changes:**
- Add `claude_common_config_snippet` field to `MultiAppConfig` struct
- Add `get_claude_common_config_snippet` and `set_claude_common_config_snippet` Tauri commands
- Include JSON validation in the setter command

**Frontend Changes:**
- Create new `lib/api/config.ts` API module
- Refactor `useCommonConfigSnippet` hook to use config.json instead of localStorage
- Add automatic one-time migration from localStorage to config.json
- Add loading state during initialization

**Benefits:**
- Cross-device synchronization via backup/restore
- More reliable persistence than browser storage
- Centralized configuration management
- Seamless migration for existing users
2025-11-13 22:45:58 +08:00
Jason 434c64f38d fix(ui): prevent URL overflow in provider cards
Add max-width constraint to URL display to prevent long URLs from breaking card layout.
2025-11-13 17:14:49 +08:00
Jason 6d8e822f8d refactor(i18n): remove unnecessary translation for brand names
- Use hardcoded "Claude", "Codex", "Gemini" instead of i18n keys
- Brand names should not be translated across different locales
- Simplifies code by removing useTranslation hook from AppSwitcher
- Reduces maintenance overhead in translation files
2025-11-13 17:08:05 +08:00
Jason 2fae8c9275 fix(i18n): add internationalization support for app names
- Add i18n for Claude/Codex/Gemini app names in AppSwitcher
- Use useTranslation hook with existing translation keys
- Fix ASCII diagram alignment in README files
2025-11-13 16:37:58 +08:00
Jason 30c763ffe3 fix(prompt): improve i18n and error handling for auto-import
Address code review feedback for PR #214:

- Replace hardcoded Chinese strings with English in auto-imported prompts
  - Prompt name: "Auto-imported Prompt" instead of "初始提示词"
  - Description: "Automatically imported on first launch"
- Remove panic risk by replacing expect() with proper error propagation
- Use AppError::localized for bilingual error messages
- Extract get_base_dir_with_fallback() helper to eliminate code duplication
- Update test assertions to match new English strings
- Suppress false-positive dead_code warning on TempHome.dir field

All 5 tests passing with zero compiler warnings.
2025-11-13 15:43:37 +08:00
YoVinchen e4d7999294 refactor(prompt): extract file path logic and implement auto-import on first launch (#214)
- Extract prompt file path logic to dedicated prompt_files module
- Refactor PromptService to use centralized path resolution
- Implement auto-import for existing prompt files on first startup
- Add comprehensive unit tests for auto-import functionality
2025-11-13 15:15:58 +08:00
YoVinchen 34f7139fda fix(usage-script): add input validation and boundary checks (#208)
- Backend: validate auto-query interval ≤ 1440 minutes (24 hours)
- Frontend: add number input sanitization and blur validation
- Add user-friendly error messages for invalid inputs
- Support auto-clamping to valid ranges with toast notifications
2025-11-13 11:28:48 +08:00
YoVinchen a85f24f616 fix(gemini): relax validation when adding providers (#210)
Allow users to create Gemini provider configurations without API key
and fill it in later. Split validation into two modes:

- validate_gemini_settings: Basic structure check (used when adding)
- validate_gemini_settings_strict: Full validation (used when switching)

This fixes the error 'Gemini config missing required field: GEMINI_API_KEY'
when trying to add Gemini providers from presets like PackyCode or Google.

Changes:
- Add validate_gemini_settings_strict for switching validation
- Update write_gemini_live to use strict validation
- Add comprehensive tests for both validation modes
2025-11-13 11:28:28 +08:00
YoVinchen b9743a463d feat(tray): add Gemini support to system tray menu (#209)
Refactor tray menu system to support three applications (Claude/Codex/Gemini):
- Introduce generic TrayAppSection structure and TRAY_SECTIONS array
- Implement append_provider_section and handle_provider_tray_event helper functions
- Enhance Gemini provider service with .env config read/write support
- Implement Gemini LiveSnapshot for atomic operations and rollback
- Update README documentation to reflect Gemini tray quick switching feature
2025-11-12 23:38:43 +08:00
YoVinchen 2f02514a14 feat(gemini): add Google Official branding with Gemini icon (#211)
Update Google Gemini preset to match Claude Official styling:
- Rename 'Google' to 'Google Official'
- Add GeminiIcon support in preset selector
- Add custom theme with Google blue (#4285F4) background
- Update PresetTheme type to support 'gemini' icon type

Changes:
- Add GeminiPresetTheme interface
- Add theme config to Google Official preset
- Import and render GeminiIcon in ProviderPresetSelector
- Update PresetTheme icon type to include 'gemini'
2025-11-12 22:41:26 +08:00
Jason 75866044bd update readme 2025-11-12 21:52:47 +08:00
YoVinchen 155532ea8c feat(prompts+i18n): add prompt management and improve prompt editor i18n (#193)
* feat(prompts): add prompt management across Tauri service and React UI

- backend: add commands/prompt.rs, services/prompt.rs, register in commands/mod.rs and lib.rs, refine app_config.rs
- frontend: add PromptPanel, PromptFormModal, PromptListItem, MarkdownEditor, usePromptActions, integrate in App.tsx
- api: add src/lib/api/prompts.ts
- i18n: update src/i18n/locales/{en,zh}.json
- build: update package.json and pnpm-lock.yaml

* feat(i18n): improve i18n for prompts and Markdown editor

- update src/i18n/locales/{en,zh}.json keys and strings
- apply i18n in PromptFormModal, PromptPanel, and MarkdownEditor
- align prompt text with src-tauri/src/services/prompt.rs

* feat(prompts): add enable/disable toggle and simplify panel UI

- Add PromptToggle component and integrate in prompt list items
- Implement toggleEnabled with optimistic update; enable via API, disable via upsert with enabled=false;
  reload after success
- Simplify PromptPanel: remove file import and current-file preview to keep CRUD flow focused
- Tweak header controls style (use mcp variant) and minor copy: rename “Prompt Management” to “Prompts”
- i18n: add disableSuccess/disableFailed messages
- Backend (Tauri): prevent duplicate backups when importing original prompt content

* style: unify code formatting with trailing commas

* feat(prompts): add Gemini filename support to PromptFormModal

Update filename mapping to use Record<AppId, string> pattern, supporting
GEMINI.md alongside CLAUDE.md and AGENTS.md.

* fix(prompts): sync enabled prompt to file when updating

When updating a prompt that is currently enabled, automatically sync
the updated content to the corresponding live file (CLAUDE.md/AGENTS.md/GEMINI.md).

This ensures the active prompt file always reflects the latest content
when editing enabled prompts.
2025-11-12 16:41:41 +08:00
YoVinchen 346f916048 refactor(endpoint): separate edit and create mode endpoint management (#192)
Optimize custom endpoint management logic to distinguish between edit and create modes:
- Edit mode: endpoints are read/written directly to backend via API
- Create mode: use draftCustomEndpoints to stage, save on submit
- Remove duplicate endpoint loading in useSpeedTestEndpoints
- Add isSaving state and initialCustomUrls tracking
2025-11-12 11:02:43 +08:00
YoVinchen 8a05e7bd3d feat(gemini): add Gemini provider integration (#202)
* feat(gemini): add Gemini provider integration

- Add gemini_config.rs module for .env file parsing
- Extend AppType enum to support Gemini
- Implement GeminiConfigEditor and GeminiFormFields components
- Add GeminiIcon with standardized 1024x1024 viewBox
- Add Gemini provider presets configuration
- Update i18n translations for Gemini support
- Extend ProviderService and McpService for Gemini

* fix(gemini): resolve TypeScript errors, add i18n support, and fix MCP logic

**Critical Fixes:**
- Fix TS2741 errors in tests/msw/state.ts by adding missing Gemini type definitions
- Fix ProviderCard.extractApiUrl to support GOOGLE_GEMINI_BASE_URL display
- Add missing apps.gemini i18n keys (zh/en) for proper app name display
- Fix MCP service Gemini cross-app duplication logic to prevent self-copy

**Technical Details:**
- tests/msw/state.ts: Add gemini default providers, current ID, and MCP config
- ProviderCard.tsx: Check both ANTHROPIC_BASE_URL and GOOGLE_GEMINI_BASE_URL
- services/mcp.rs: Skip Gemini in sync_other_side logic with unreachable!() guards
- Run pnpm format to auto-fix code style issues

**Verification:**
-  pnpm typecheck passes
-  pnpm format completed

* feat(gemini): enhance authentication and config parsing

- Add strict and lenient .env parsing modes
- Implement PackyCode partner authentication detection
- Support Google OAuth official authentication
- Auto-configure security.auth.selectedType for PackyCode
- Add comprehensive test coverage for all auth types
- Update i18n for OAuth hints and Gemini config

---------

Co-authored-by: Jason <farion1231@gmail.com>
2025-11-12 10:47:34 +08:00
Jason 32a2ba5ef6 chore(release): prepare for v3.6.2 release 2025-11-11 23:57:21 +08:00
Jason 4502b2f973 feat(presets): add Kimi For Coding and BaiLing provider presets
Add two new Chinese official provider presets:
- Kimi For Coding: AI coding assistant from Kimi
- BaiLing: Claude-compatible API from AliPay TBox
2025-11-11 23:50:52 +08:00
Jason 6cb930b4ec docs: add TypeScript Trending badge and improve contributing tone
- Add 🔥 TypeScript Trending badge celebrating daily/weekly/monthly rankings
- Refine contributing guidelines with friendlier, more welcoming language
- Replace directive tone with collaborative suggestions for feature PRs
2025-11-11 11:03:26 +08:00
Jason 9a5c8c0e57 chore(release): prepare for v3.6.1 release
- Bump version to 3.6.1 across all config files
  - package.json, Cargo.toml, tauri.conf.json
  - README.md and README_ZH.md version badges
  - Auto-updated Cargo.lock
- Add release notes documentation
  - docs/release-note-v3.6.0-en.md (archive)
  - docs/release-note-v3.6.1-zh.md (cumulative format)
  - docs/release-note-v3.6.1-en.md (cumulative format)
- Enlarge PackyCode sponsor logo by 50% (100px → 150px)
- Update roadmap checklist (homebrew support marked as done)
2025-11-10 21:21:27 +08:00
Jason b1abdf95aa feat(sponsor): add PackyCode as official partner
- Add PackyCode to sponsor section in README (both EN/ZH)
- Add PackyCode logo to assets/partners/logos/
- Mark PackyCode as partner in provider presets (Claude & Codex)
- Add promotion message to i18n files with 10% discount info
2025-11-10 18:44:19 +08:00
Jason be155c857e refactor(usage-script): replace native checkbox with Switch component
Upgrade the enable toggle from native checkbox to shadcn/ui Switch component
for better UX and UI consistency with settings page.

**Improvements**:
1. Use modern toggle UI (Switch) instead of traditional checkbox
2. Adopt the same layout pattern as settings page (ToggleRow style)
3. Add bordered container with proper spacing for better visual hierarchy
4. Maintain full accessibility support (aria-label)

**Layout changes**:
- Before: Simple label + checkbox horizontal layout
- After: Bordered container, label on left, Switch on right, vertically centered
2025-11-10 16:01:28 +08:00
Jason 9d6f101006 fix(usage-script): replace FormLabel with Label to fix white screen crash
FormLabel component requires FormField context and throws error when used
standalone, causing the entire component to crash with a white screen.

Root cause:
- FormLabel internally calls useFormField() hook
- useFormField() requires FormFieldContext (must be within <FormField>)
- Without context, it throws: "useFormField should be used within <FormField>"
- Uncaught error crashes React rendering tree

Solution:
- Replace FormLabel with standalone Label component
- Label component from @/components/ui/label doesn't depend on form context
- Maintains same styling (text-sm font-medium) without requiring context

This fixes the white screen issue when clicking the usage panel.
2025-11-10 15:51:18 +08:00
Jason 2a56a0d889 style(usage-script): unify form input styles with shadcn/ui components
Replace native HTML input elements with shadcn/ui Input and FormLabel
components to ensure consistent styling across the application.

Changes:
- Import Input, FormLabel, Eye, and EyeOff components
- Replace all credential input fields with Input component
- Add show/hide toggle buttons for password fields (API Key, Access Token)
- Replace label/span elements with FormLabel component
- Update timeout and auto-query interval inputs to use Input component
- Improve spacing consistency (space-y-4 for credential config)
- Add proper id attributes for accessibility
- Use muted-foreground for hint text

The form now matches the styling of provider configuration forms
throughout the application.
2025-11-10 15:42:36 +08:00
Jason 7096957b40 feat(usage-query): decouple credentials from provider config
Add independent credential fields for usage query to support different
query endpoints and authentication methods.

Changes:
- Add `apiKey` and `baseUrl` fields to UsageScript struct
- Remove dependency on provider config credentials in query_usage
- Update test_usage_script to accept independent credential parameters
- Add credential input fields in UsageScriptModal based on template:
  * General: apiKey + baseUrl
  * NewAPI: baseUrl + accessToken + userId
  * Custom: no additional fields (full freedom)
- Auto-clear irrelevant fields when switching templates
- Add i18n text for "credentialsConfig"

Benefits:
- Query API can use different endpoint/key than provider config
- Better separation of concerns
- More flexible for various usage query scenarios
2025-11-10 15:28:09 +08:00
Jason 23d06515ad fix(toml): normalize CJK quotes to prevent parsing errors
Add quote normalization to handle Chinese/fullwidth quotes automatically
converted by IME. This fixes TOML parsing failures when users input
configuration with non-ASCII quotes (" " ' ' etc.).

Changes:
- Add textNormalization utility for quote normalization
- Apply normalization in TOML input handlers (MCP form, Codex config)
- Disable browser auto-correction in Textarea component
- Add defensive normalization in TOML parsing layer
2025-11-10 14:35:55 +08:00
Jason 3210202132 fix(mcp): preserve custom fields in Codex TOML config editor
Fixed an issue where custom/extension fields (e.g., timeout_ms, retry_count)
were silently dropped when editing Codex MCP server configurations in TOML format.

Root cause: The TOML parser functions only extracted known fields (type, command,
args, env, cwd, url, headers), discarding any additional fields during normalization.

Changes:
- mcpServerToToml: Now uses spread operator to copy all fields before stringification
- normalizeServerConfig: Added logic to preserve unknown fields after processing known ones
- Both stdio and http server types now retain custom configuration fields

This fix enables forward compatibility with future MCP protocol extensions and
allows users to add custom configurations without code changes.
2025-11-10 12:03:15 +08:00
Jason 7b52c44a9d feat(schema): add common JSON/TOML validators and enforce MCP conditional fields
- Add src/lib/schemas/common.ts with jsonConfigSchema and tomlConfigSchema
- Enhance src/lib/schemas/mcp.ts to require command for stdio and url for http via superRefine
- Keep ProviderForm as-is; future steps will wire new schemas into RHF flows
- Verified: pnpm typecheck passes
2025-11-09 20:42:43 +08:00
Jason 772081312e feat(ui): enhance provider switch error notification with copy action
- Split error message into title and description for better UX
- Add one-click copy button to easily share error details
- Extend toast duration to 6s for improved readability
- Use extractErrorMessage utility for consistent error handling
- Add i18n keys: common.copy, notifications.switchFailedTitle

This improves debugging experience when provider switching fails,
allowing users to quickly copy and share error messages.
2025-11-09 17:56:02 +08:00
Jason cfcd7b892a fix(tray): replace unwrap with safe pattern matching in menu handler
Replace unwrap() calls with safe pattern matching to prevent panics
when handling invalid tray menu item IDs. Now logs errors and returns
gracefully instead of crashing the application.
2025-11-08 22:55:53 +08:00
Jason 3da787b9af fix(provider): set category to custom for imported default config
Ensure that when importing default configuration from live files,
the created provider is properly marked with category "custom" for
correct UI display and filtering.
2025-11-08 22:45:53 +08:00
Jason 9370054911 fix(error-handling): isolate tray menu update failures from main operations
Previously, if updateTrayMenu() failed after a successful main operation
(like sorting, adding, or updating providers), the entire operation would
appear to fail with a misleading error message, even though the core
functionality had already succeeded.

This resulted in false negative feedback where:
- Backend data was successfully updated
- Frontend UI was successfully refreshed
- Tray menu failed to update
- User saw "operation failed" message (incorrect)

Changes:
- Wrap updateTrayMenu() calls in nested try-catch blocks
- Log tray menu failures separately with descriptive messages
- Ensure main operation success is reported accurately
- Prevent tray menu failures from triggering main operation error handlers

Files modified:
- src/hooks/useDragSort.ts (drag-and-drop sorting)
- src/lib/query/mutations.ts (add/delete/switch mutations)
- src/hooks/useProviderActions.ts (update provider)

This fixes the bug introduced in PR #179 and prevents similar issues
across all provider operations.
2025-11-08 22:07:12 +08:00
ZyphrZero 5b3b211c9a fix(ui): sync tray menu order after drag-and-drop sorting (#179)
The drag-and-drop sorting feature introduced in PR #126 (9eb991d) was
updating the provider sort order in the backend and frontend, but failed
to update the tray menu to reflect the new order.

Changes:
- Add updateTrayMenu() call after successful sort order update
- Ensures tray menu items are immediately reordered to match the UI

This fixes the issue where dragging providers in the main window would
not update their order in the system tray menu.

Fixes: 9eb991d (feat(ui): add drag-and-drop sorting for provider list)
2025-11-08 21:45:43 +08:00
Jason fb02881684 fix(forms): populate base URL for all non-official provider categories
The base URL field was not populating when editing providers with
categories like cn_official or aggregator. The issue was caused by
inconsistent conditional logic: the input field was shown for all
non-official categories, but the value extraction only worked for
third_party and custom categories.

Changed the category check from allowlist (third_party, custom) to
denylist (official) to match the UI display logic. Now ANTHROPIC_BASE_URL
correctly populates for all provider categories except official.
2025-11-08 21:25:41 +08:00
Jason 34b8aa1008 fix(ui): remove misleading model placeholders from input fields
Clear placeholder values for model input fields to avoid suggesting specific model names that may not be applicable to all providers. This prevents user confusion when configuring different Claude providers.
2025-11-08 08:50:15 +08:00
Jason 52a7f9d313 feat(usage): enable background query for usage polling
- Add `refetchIntervalInBackground: true` to usage query configuration
- Allows usage queries to continue running when app window is minimized or unfocused
- Existing safety mechanisms remain in place (timeout limits, minimum interval, no retry)
- Users have full control through autoQueryInterval setting
2025-11-08 00:02:28 +08:00
Jason b617879035 docs: add v3.6.0 release notes and update READMEs with missing features
**Release Notes**
- Create Chinese v3.6.0 release notes in docs/ folder

**README Updates**
Add documentation for three missing v3.6.0 features:

1. Claude Configuration Data Structure Enhancements
   - Granular model configuration: migrated from dual-key to quad-key system
   - New fields: ANTHROPIC_DEFAULT_HAIKU_MODEL, ANTHROPIC_DEFAULT_SONNET_MODEL,
     ANTHROPIC_DEFAULT_OPUS_MODEL, ANTHROPIC_MODEL
   - Replaces legacy ANTHROPIC_SMALL_FAST_MODEL with automatic migration
   - Backend normalizes old configs with smart fallback chain
   - UI expanded from 2 to 4 model input fields

2. Updated Provider Models
   - Kimi: updated to latest kimi-k2-thinking model (from k1 series)
   - Removed legacy KimiModelSelector component

3. Custom Configuration Directory (Cloud Sync Support)
   - Customize CC Switch's configuration storage location
   - Point to cloud sync folders (Dropbox, OneDrive, iCloud, etc.)
     to enable automatic config synchronization across devices
   - Managed via Tauri Store for better isolation

**Files Changed**
- README.md & README_ZH.md: added feature documentation
- docs/release-note-v3.6.0-zh.md: comprehensive Chinese release notes
2025-11-07 22:05:33 +08:00
367 changed files with 51209 additions and 8325 deletions
+6
View File
@@ -9,6 +9,12 @@ release/
.npmrc
CLAUDE.md
AGENTS.md
GEMINI.md
/.claude
/.codex
/.gemini
/.cc-switch
/.idea
/.vscode
vitest-report.json
nul
+287
View File
@@ -5,6 +5,287 @@ All notable changes to CC Switch will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [3.8.0] - 2025-11-28
### Major Updates
- **Persistence architecture upgrade** - Moved from single JSON storage to SQLite + JSON dual-layer; added schema versioning, transactions, and SQL import/export; first launch auto-migrates `config.json` to SQLite while keeping originals safe.
- **Brand new UI** - Full layout redesign, unified component/ConfirmDialog styles, smoother animations, overscroll disabled; Tailwind CSS downgraded to v3.4 for compatibility.
- **Japanese language support** - UI now localized in Chinese/English/Japanese.
### Added
- **Skills recursive scanning** - Discovers nested `SKILL.md` files across multi-level directories; same-name skills allowed by full-path dedup.
- **Provider icons** - Presets ship with default icons; custom icon colors; icons retained when duplicating providers.
- **Auto launch on startup** - One-click enable/disable using Registry/LaunchAgent/XDG autostart.
- **Provider preset** - Added MiniMax partner preset.
- **Form validation** - Required fields get real-time validation and unified toast messaging.
### Fixed
- **Custom endpoints loss** - Switched provider updates to `UPDATE` to avoid cascade deletes from `INSERT OR REPLACE`.
- **Gemini config writing** - Correctly writes custom env vars to `.env` and keeps auth configs isolated.
- **Provider validation** - Handles missing current provider IDs and preserves icon fields on duplicate.
- **Linux rendering** - Fixed WebKitGTK DMA-BUF rendering and preserved user `.desktop` customizations.
- **Misc** - Removed redundant usage queries; corrected DMXAPI auth token field; restored missing deeplink translations; fixed usage script template init.
### Technical
- **Database modules** - Added `schema`, `backup`, `migration`, and DAO layers for providers/MCP/prompts/skills/settings.
- **Service modularization** - Split provider service into live/auth/endpoints/usage modules; deeplink parsing/import logic modularized.
- **Code cleanup** - Removed legacy JSON-era import/export, unused MCP types; unified error handling; tests migrated to SQLite backend and MSW handlers updated.
### Migration Notes
- First launch auto-migrates data from `config.json` to SQLite and device settings to `settings.json`; originals kept; error dialog on failure; dry-run supported.
### Stats
- 51 commits since v3.7.1; 207 files changed; +17,297 / -6,870 lines. See [release-note-v3.8.0](docs/release-note-v3.8.0-en.md) for details.
---
## [3.7.1] - 2025-11-22
### Fixed
- **Skills third-party repository installation** (#268) - Fixed installation failure for skills repositories with custom subdirectories (e.g., `ComposioHQ/awesome-claude-skills`)
- **Gemini configuration persistence** - Resolved issue where settings.json edits were lost when switching providers
- **Dialog overlay click protection** - Prevented dialogs from closing when clicking outside, avoiding accidental form data loss (affects 11 dialog components)
### Added
- **Gemini configuration directory support** (#255) - Added custom configuration directory option for Gemini in settings
- **ArchLinux installation support** (#259) - Added AUR installation via `paru -S cc-switch-bin`
### Improved
- **Skills error messages i18n** - Added 28+ detailed error messages (English & Chinese) with specific resolution suggestions
- **Download timeout** - Extended from 15s to 60s to reduce network-related false positives
- **Code formatting** - Applied unified Rust (`cargo fmt`) and TypeScript (`prettier`) formatting standards
### Reverted
- **Auto-launch on system startup** - Temporarily reverted feature pending further testing and optimization
---
## [3.7.0] - 2025-11-19
### Major Features
#### Gemini CLI Integration
- **Complete Gemini CLI support** - Third major application added alongside Claude Code and Codex
- **Dual-file configuration** - Support for both `.env` and `settings.json` file formats
- **Environment variable detection** - Auto-detect `GOOGLE_GEMINI_BASE_URL`, `GEMINI_MODEL`, etc.
- **MCP management** - Full MCP configuration capabilities for Gemini
- **Provider presets**
- Google Official (OAuth authentication)
- PackyCode (partner integration)
- Custom endpoint support
- **Deep link support** - Import Gemini providers via `ccswitch://` protocol
- **System tray integration** - Quick-switch Gemini providers from tray menu
- **Backend modules** - New `gemini_config.rs` (20KB) and `gemini_mcp.rs`
#### MCP v3.7.0 Unified Architecture
- **Unified management panel** - Single interface for Claude/Codex/Gemini MCP servers
- **SSE transport type** - New Server-Sent Events support alongside stdio/http
- **Smart JSON parser** - Fault-tolerant parsing of various MCP config formats
- **Extended field support** - Preserve custom fields in Codex TOML conversion
- **Codex format correction** - Proper `[mcp_servers]` format (auto-cleanup of incorrect `[mcp.servers]`)
- **Import/export system** - Unified import from Claude/Codex/Gemini live configs
- **UX improvements**
- Default app selection in forms
- JSON formatter for config validation
- Improved layout and visual hierarchy
- Better validation error messages
#### Claude Skills Management System
- **GitHub repository integration** - Auto-scan and discover skills from GitHub repos
- **Pre-configured repositories**
- `ComposioHQ/awesome-claude-skills` (curated collection)
- `anthropics/skills` (official Anthropic skills)
- `cexll/myclaude` (community, with subdirectory scanning)
- **Lifecycle management**
- One-click install to `~/.claude/skills/`
- Safe uninstall with state tracking
- Update checking (infrastructure ready)
- **Custom repository support** - Add any GitHub repo as a skill source
- **Subdirectory scanning** - Optional `skillsPath` for repos with nested skill directories
- **Backend architecture** - `SkillService` (526 lines) with GitHub API integration
- **Frontend interface**
- SkillsPage: Browse and manage skills
- SkillCard: Visual skill presentation
- RepoManager: Repository management dialog
- **State persistence** - Installation state stored in `skills.json`
- **Full i18n support** - Complete Chinese/English translations (47+ keys)
#### Prompts (System Prompts) Management
- **Multi-preset management** - Create, edit, and switch between multiple system prompts
- **Cross-app support**
- Claude: `~/.claude/CLAUDE.md`
- Codex: `~/.codex/AGENTS.md`
- Gemini: `~/.gemini/GEMINI.md`
- **Markdown editor** - Full-featured CodeMirror 6 editor with syntax highlighting
- **Smart synchronization**
- Auto-write to live files on enable
- Content backfill protection (save current before switching)
- First-launch auto-import from live files
- **Single-active enforcement** - Only one prompt can be active at a time
- **Delete protection** - Cannot delete active prompts
- **Backend service** - `PromptService` (213 lines) with CRUD operations
- **Frontend components**
- PromptPanel: Main management interface (177 lines)
- PromptFormModal: Edit dialog with validation (160 lines)
- MarkdownEditor: CodeMirror integration (159 lines)
- usePromptActions: Business logic hook (152 lines)
- **Full i18n support** - Complete Chinese/English translations (41+ keys)
#### Deep Link Protocol (ccswitch://)
- **Protocol registration** - `ccswitch://` URL scheme for one-click imports
- **Provider import** - Import provider configurations from URLs or shared links
- **Lifecycle integration** - Deep link handling integrated into app startup
- **Cross-platform support** - Works on Windows, macOS, and Linux
#### Environment Variable Conflict Detection
- **Claude & Codex detection** - Identify conflicting environment variables
- **Gemini auto-detection** - Automatic environment variable discovery
- **Conflict management** - UI for resolving configuration conflicts
- **Prevention system** - Warn before overwriting existing configurations
### New Features
#### Provider Management
- **DouBaoSeed preset** - Added ByteDance's DouBao provider
- **Kimi For Coding** - Moonshot AI coding assistant
- **BaiLing preset** - BaiLing AI integration
- **Removed AnyRouter preset** - Discontinued provider
- **Model configuration** - Support for custom model names in Codex and Gemini
- **Provider notes field** - Add custom notes to providers for better organization
#### Configuration Management
- **Common config migration** - Moved Claude common config snippets from localStorage to `config.json`
- **Unified persistence** - Common config snippets now shared across all apps
- **Auto-import on first launch** - Automatically import configs from live files on first run
- **Backfill priority fix** - Correct priority handling when enabling prompts
#### UI/UX Improvements
- **macOS native design** - Migrated color scheme to macOS native design system
- **Window centering** - Default window position centered on screen
- **Password input fixes** - Disabled Edge/IE reveal and clear buttons
- **URL overflow prevention** - Fixed overflow in provider cards
- **Error notification enhancement** - Copy-to-clipboard for error messages
- **Tray menu sync** - Real-time sync after drag-and-drop sorting
### Improvements
#### Architecture
- **MCP v3.7.0 cleanup** - Removed legacy code and warnings
- **Unified structure** - Default initialization with v3.7.0 unified structure
- **Backward compatibility** - Compilation fixes for older configs
- **Code formatting** - Applied consistent formatting across backend and frontend
#### Platform Compatibility
- **Windows fix** - Resolved winreg API compatibility issue (v0.52)
- **Safe pattern matching** - Replaced `unwrap()` with safe patterns in tray menu
#### Configuration
- **MCP sync on switch** - Sync MCP configs for all apps when switching providers
- **Gemini form sync** - Fixed form fields syncing with environment editor
- **Gemini config reading** - Read from both `.env` and `settings.json`
- **Validation improvements** - Enhanced input validation and boundary checks
#### Internationalization
- **JSON syntax fixes** - Resolved syntax errors in locale files
- **App name i18n** - Added internationalization support for app names
- **Deduplicated labels** - Reused providerForm keys to reduce duplication
- **Gemini MCP title** - Added missing Gemini MCP panel title
### Bug Fixes
#### Critical Fixes
- **Usage script validation** - Added input validation and boundary checks
- **Gemini validation** - Relaxed validation when adding providers
- **TOML quote normalization** - Handle CJK quotes to prevent parsing errors
- **MCP field preservation** - Preserve custom fields in Codex TOML editor
- **Password input** - Fixed white screen crash (FormLabel → Label)
#### Stability
- **Tray menu safety** - Replaced unwrap with safe pattern matching
- **Error isolation** - Tray menu update failures don't block main operations
- **Import classification** - Set category to custom for imported default configs
#### UI Fixes
- **Model placeholders** - Removed misleading model input placeholders
- **Base URL population** - Auto-fill base URL for non-official providers
- **Drag sort sync** - Fixed tray menu order after drag-and-drop
### Technical Improvements
#### Code Quality
- **Type safety** - Complete TypeScript type coverage across codebase
- **Test improvements** - Simplified boolean assertions in tests
- **Clippy warnings** - Fixed `uninlined_format_args` warnings
- **Code refactoring** - Extracted templates, optimized logic flows
#### Dependencies
- **Tauri** - Updated to 2.8.x series
- **Rust dependencies** - Added `anyhow`, `zip`, `serde_yaml`, `tempfile` for Skills
- **Frontend dependencies** - Added CodeMirror 6 packages for Markdown editor
- **winreg** - Updated to v0.52 (Windows compatibility)
#### Performance
- **Startup optimization** - Removed legacy migration scanning
- **Lock management** - Improved RwLock usage to prevent deadlocks
- **Background query** - Enabled background mode for usage polling
### Statistics
- **Total commits**: 85 commits from v3.6.0 to v3.7.0
- **Code changes**: 152 files changed, 18,104 insertions(+), 3,732 deletions(-)
- **New modules**:
- Skills: 2,034 lines (21 files)
- Prompts: 1,302 lines (20 files)
- Gemini: ~1,000 lines (multiple files)
- MCP refactor: ~3,000 lines (refactored)
### Strategic Positioning
v3.7.0 represents a major evolution from "Provider Switcher" to **"All-in-One AI CLI Management Platform"**:
1. **Capability Extension** - Skills provide external ability integration
2. **Behavior Customization** - Prompts enable AI personality presets
3. **Configuration Unification** - MCP v3.7.0 eliminates app silos
4. **Ecosystem Openness** - Deep links enable community sharing
5. **Multi-AI Support** - Claude/Codex/Gemini trinity
6. **Intelligent Detection** - Auto-discovery of environment conflicts
### Notes
- Users upgrading from v3.1.0 or earlier should first upgrade to v3.2.x for one-time migration
- Skills and Prompts management are new features requiring no migration
- Gemini CLI support requires Gemini CLI to be installed separately
- MCP v3.7.0 unified structure is backward compatible with previous configs
## [3.6.0] - 2025-11-07
### ✨ New Features
@@ -73,6 +354,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
### 🏗️ Technical Improvements (For Developers)
**Backend Refactoring (Rust)** - Completed 5-phase refactoring:
- **Phase 1**: Unified error handling (`AppError` + i18n error messages)
- **Phase 2**: Command layer split by domain (`commands/{provider,mcp,config,settings,plugin,misc}.rs`)
- **Phase 3**: Integration tests and transaction mechanism (config snapshot + failure rollback)
@@ -80,17 +362,20 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **Phase 5**: Concurrency optimization (`RwLock` instead of `Mutex`, scoped guard to avoid deadlock)
**Frontend Refactoring (React + TypeScript)** - Completed 4-stage refactoring:
- **Stage 1**: Test infrastructure (vitest + MSW + @testing-library/react)
- **Stage 2**: Extracted custom hooks (`useProviderActions`, `useMcpActions`, `useSettings`, `useImportExport`, etc.)
- **Stage 3**: Component splitting and business logic extraction
- **Stage 4**: Code cleanup and formatting unification
**Testing System**:
- Hooks unit tests 100% coverage
- Integration tests covering key processes (App, SettingsDialog, MCP Panel)
- MSW mocking backend API to ensure test independence
**Code Quality**:
- Unified parameter format: All Tauri commands migrated to camelCase (Tauri 2 specification)
- `AppType` renamed to `AppId`: Semantically clearer
- Unified parsing with `FromStr` trait: Centralized `app` parameter parsing
@@ -98,6 +383,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- Remove unused code: `missing_param` helper function, deprecated `tauri-api.ts`, redundant `KimiModelSelector` component
**Internal Optimizations**:
- **Removed Legacy Migration Logic**: v3.6 removed v1 config auto-migration and copy file scanning logic
-**Impact**: Improved startup performance, cleaner code
-**Compatibility**: v2 format configs fully compatible, no action required
@@ -361,6 +647,7 @@ For users upgrading from v2.x (Electron version):
- Basic provider management
- Claude Code integration
- Configuration file handling
## [Unreleased]
### ⚠️ Breaking Changes
+249 -290
View File
@@ -1,18 +1,24 @@
# Claude Code & Codex Provider Switcher
<div align="center">
[![Version](https://img.shields.io/badge/version-3.6.0-blue.svg)](https://github.com/farion1231/cc-switch/releases)
# All-in-One Assistant for Claude Code, Codex & Gemini CLI
[![Version](https://img.shields.io/badge/version-3.8.2-blue.svg)](https://github.com/farion1231/cc-switch/releases)
[![Trending](https://img.shields.io/badge/🔥_TypeScript_Trending-Daily%20%7C%20Weekly%20%7C%20Monthly-ff6b6b.svg)](https://github.com/trending/typescript)
[![Platform](https://img.shields.io/badge/platform-Windows%20%7C%20macOS%20%7C%20Linux-lightgrey.svg)](https://github.com/farion1231/cc-switch/releases)
[![Built with Tauri](https://img.shields.io/badge/built%20with-Tauri%202-orange.svg)](https://tauri.app/)
[![Downloads](https://img.shields.io/endpoint?url=https://api.pinstudios.net/api/badges/downloads/farion1231/cc-switch/total)](https://github.com/farion1231/cc-switch/releases/latest)
English | [中文](README_ZH.md) | [Changelog](CHANGELOG.md)
<a href="https://trendshift.io/repositories/15372" target="_blank"><img src="https://trendshift.io/api/badge/repositories/15372" alt="farion1231%2Fcc-switch | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
A desktop application for managing and switching between different provider configurations & MCP for Claude Code and Codex.
English | [中文](README_ZH.md) | [日本語](README_JA.md) | [Changelog](CHANGELOG.md)
**From Provider Switcher to All-in-One AI CLI Management Platform**
Unified management for Claude Code, Codex & Gemini CLI provider configurations, MCP servers, Skills extensions, and system prompts.
</div>
## ❤️ Sponsor
## ❤️Sponsor
![Zhipu GLM](assets/partners/banners/glm-en.jpg)
@@ -22,78 +28,124 @@ GLM CODING PLAN is a subscription service designed for AI coding, starting at ju
Get 10% OFF the GLM CODING PLAN with [this link](https://z.ai/subscribe?ic=8JVLJQFSKB)!
## Release Notes
---
> **v3.6.0**: Added edit mode (provider duplication, manual sorting), custom endpoint management, usage query features. Optimized config directory switching experience (perfect WSL environment support). Added multiple provider presets (DMXAPI, Azure Codex, AnyRouter, AiHubMix, MiniMax). Completed full-stack architecture refactoring and testing infrastructure.
<table>
<tr>
<td width="180"><img src="assets/partners/logos/packycode.png" alt="PackyCode" width="150"></td>
<td>Thanks to PackyCode for sponsoring this project! PackyCode is a reliable and efficient API relay service provider, offering relay services for Claude Code, Codex, Gemini, and more. PackyCode provides special discounts for our software users: register using <a href="https://www.packyapi.com/register?aff=cc-switch">this link</a> and enter the "cc-switch" promo code during recharge to get 10% off.</td>
</tr>
> v3.5.0: Added MCP management, config import/export, endpoint speed testing. Complete i18n coverage. Added Longcat and kat-coder presets. Standardized release file naming conventions.
<tr>
<td width="180"><img src="assets/partners/logos/sds-en.png" alt="ShanDianShuo" width="150"></td>
<td>Thanks to ShanDianShuo for sponsoring this project! ShanDianShuo is a local-first AI voice input: Millisecond latency, data stays on device, 4x faster than typing, AI-powered correction, Privacy-first, completely free. Doubles your coding efficiency with Claude Code! <a href="https://www.shandianshuo.cn">Free download</a> for Mac/Win</td>
</tr>
> v3.4.0: Added i18next internationalization, support for new models (qwen-3-max, GLM-4.6, DeepSeek-V3.2-Exp), Claude plugin, single-instance daemon, tray minimize, and installer optimizations.
> v3.3.0: One-click VS Code Codex plugin configuration/removal (auto-sync by default), Codex common config snippets, enhanced custom wizard, WSL environment support, cross-platform tray and UI optimizations. (VS Code write feature deprecated in v3.4.x)
> v3.2.0: Brand new UI, macOS system tray, built-in updater, atomic write with rollback, improved dark mode, Single Source of Truth (SSOT) with one-time migration/archival.
> v3.1.0: Added Codex provider management with one-click switching. Import current Codex config as default provider. Auto-backup before internal config v1 → v2 migration (see "Migration & Archival" below).
> v3.0.0 Major Update: Complete migration from Electron to Tauri 2.0. Significantly reduced app size and greatly improved startup performance.
## Features (v3.6.0)
### Core Features
- **MCP (Model Context Protocol) Management**: Complete MCP server configuration management system
- Support for stdio and http server types with command validation
- Built-in templates for popular MCP servers (e.g., mcp-fetch)
- Real-time enable/disable MCP servers with atomic file writes to prevent configuration corruption
- **Config Import/Export**: Backup and restore your provider configurations
- One-click export all configurations to JSON file
- Import configs with automatic validation and backup, auto-rotate backups (keep 10 most recent)
- Auto-sync to live config files after import to ensure immediate effect
- **Endpoint Speed Testing**: Test API endpoint response times
- Measure latency to different provider endpoints with visual connection quality indicators
- Help users choose the fastest provider
- **Internationalization & Language Switching**: Complete i18next i18n coverage (including error messages, tray menu, all UI components)
- **Claude Plugin Sync**: Built-in button to apply or restore Claude plugin configurations with one click. Takes effect immediately after switching providers.
### v3.6 New Features
- **Provider Duplication**: Quickly duplicate existing provider configs to easily create variants
- **Manual Sorting**: Drag and drop to manually reorder providers
- **Custom Endpoint Management**: Support multi-endpoint configuration for aggregator providers
- **Usage Query Features**
- Auto-refresh interval: Supports periodic automatic usage queries
- Test Script API: Validate JavaScript scripts before execution
- Template system expansion: Custom blank templates, support for access token and user ID parameters
- **Config Editor Improvements**
- Added JSON format button
- Real-time TOML syntax validation (for Codex configs)
- **Auto-sync on Directory Change**: When switching Claude/Codex config directories (e.g., switching to WSL environment), automatically sync current provider to new directory to avoid config file conflicts
- **Load Live Config When Editing Active Provider**: When editing the currently active provider, prioritize displaying the actual effective configuration to protect user manual modifications
- **New Provider Presets**: DMXAPI, Azure Codex, AnyRouter, AiHubMix, MiniMax
- **Partner Promotion Mechanism**: Support ecosystem partner promotion (e.g., Zhipu GLM Z.ai)
### v3.6 Architecture Improvements
- **Backend Refactoring**: Completed 5-phase refactoring (unified error handling → command layer split → integration tests → Service layer extraction → concurrency optimization)
- **Frontend Refactoring**: Completed 4-stage refactoring (test infrastructure → Hooks extraction → component splitting → code cleanup)
- **Testing System**: 100% Hooks unit test coverage, integration tests covering critical flows (vitest + MSW + @testing-library/react)
### System Features
- **System Tray & Window Behavior**: Window can minimize to tray, macOS supports hide/show Dock in tray mode, tray switching syncs Claude/Codex/plugin status.
- **Single Instance**: Ensures only one instance runs at a time to avoid multi-instance conflicts.
- **Standardized Release Naming**: All platform release files use consistent version-tagged naming (macOS: `.tar.gz` / `.zip`, Windows: `.msi` / `-Portable.zip`, Linux: `.AppImage` / `.deb`).
</table>
## Screenshots
### Main Interface
| Main Interface | Add Provider |
| :-----------------------------------------------: | :--------------------------------------------: |
| ![Main Interface](assets/screenshots/main-en.png) | ![Add Provider](assets/screenshots/add-en.png) |
![Main Interface](assets/screenshots/main-en.png)
## Features
### Add Provider
### Current Version: v3.8.2 | [Full Changelog](CHANGELOG.md) | [Release Notes](docs/release-note-v3.8.0-en.md)
![Add Provider](assets/screenshots/add-en.png)
**v3.8.0 Major Update (2025-11-28)**
**Persistence Architecture Upgrade & Brand New UI**
- **SQLite + JSON Dual-layer Architecture**
- Migrated from JSON file storage to SQLite + JSON dual-layer structure
- Syncable data (providers, MCP, Prompts, Skills) stored in SQLite
- Device-level data (window state, local paths) stored in JSON
- Lays the foundation for future cloud sync functionality
- Schema version management for database migrations
- **Brand New User Interface**
- Completely redesigned interface layout
- Unified component styles and smoother animations
- Optimized visual hierarchy
- Tailwind CSS downgraded from v4 to v3.4 for better browser compatibility
- **Japanese Language Support**
- Added Japanese interface support (now supports Chinese/English/Japanese)
- **Auto Launch on Startup**
- One-click enable/disable in settings
- Platform-native APIs (Registry/LaunchAgent/XDG autostart)
- **Skills Recursive Scanning**
- Support for multi-level directory structures
- Allow same-named skills from different repositories
- **Critical Bug Fixes**
- Fixed custom endpoints lost when updating providers
- Fixed Gemini configuration write issues
- Fixed Linux WebKitGTK rendering issues
**v3.7.0 Highlights**
**Six Core Features, 18,000+ Lines of New Code**
- **Gemini CLI Integration**
- Third supported AI CLI (Claude Code / Codex / Gemini)
- Dual-file configuration support (`.env` + `settings.json`)
- Complete MCP server management
- Presets: Google Official (OAuth) / PackyCode / Custom
- **Claude Skills Management System**
- Auto-scan skills from GitHub repositories (3 pre-configured curated repos)
- One-click install/uninstall to `~/.claude/skills/`
- Custom repository support + subdirectory scanning
- Complete lifecycle management (discover/install/update)
- **Prompts Management System**
- Multi-preset system prompt management (unlimited presets, quick switching)
- Cross-app support (Claude: `CLAUDE.md` / Codex: `AGENTS.md` / Gemini: `GEMINI.md`)
- Markdown editor (CodeMirror 6 + real-time preview)
- Smart backfill protection, preserves manual modifications
- **MCP v3.7.0 Unified Architecture**
- Single panel manages MCP servers across three applications
- New SSE (Server-Sent Events) transport type
- Smart JSON parser + Codex TOML format auto-correction
- Unified import/export + bidirectional sync
- **Deep Link Protocol**
- `ccswitch://` protocol registration (all platforms)
- One-click import provider configs via shared links
- Security validation + lifecycle integration
- **Environment Variable Conflict Detection**
- Auto-detect cross-app configuration conflicts (Claude/Codex/Gemini/MCP)
- Visual conflict indicators + resolution suggestions
- Override warnings + backup before changes
**Core Capabilities**
- **Provider Management**: One-click switching between Claude Code, Codex, and Gemini API configurations
- **Speed Testing**: Measure API endpoint latency with visual quality indicators
- **Import/Export**: Backup and restore configs with auto-rotation (keep 10 most recent)
- **i18n Support**: Complete Chinese/English localization (UI, errors, tray)
- **Claude Plugin Sync**: One-click apply/restore Claude plugin configurations
**v3.6 Highlights**
- Provider duplication & drag-and-drop sorting
- Multi-endpoint management & custom config directory (cloud sync ready)
- Granular model configuration (4-tier: Haiku/Sonnet/Opus/Custom)
- WSL environment support with auto-sync on directory change
- 100% hooks test coverage & complete architecture refactoring
**System Features**
- System tray with quick switching
- Single instance daemon
- Built-in auto-updater
- Atomic writes with rollback protection
## Download & Installation
@@ -128,152 +180,143 @@ Download `CC-Switch-v{version}-macOS.zip` from the [Releases](../../releases) pa
> **Note**: Since the author doesn't have an Apple Developer account, you may see an "unidentified developer" warning on first launch. Please close it first, then go to "System Settings" → "Privacy & Security" → click "Open Anyway", and you'll be able to open it normally afterwards.
### ArchLinux 用户
**Install via paru (Recommended)**
```bash
paru -S cc-switch-bin
```
### Linux Users
Download the latest `CC-Switch-v{version}-Linux.deb` package or `CC-Switch-v{version}-Linux.AppImage` from the [Releases](../../releases) page.
## Usage Guide
## Quick Start
1. Click "Add Provider" to add your API configuration
2. Switching methods:
- Select a provider on the main interface and click switch
- Or directly select target provider from "System Tray (Menu Bar)" for immediate effect
3. Switching will write to the corresponding app's "live config file" (Claude: `settings.json`; Codex: `auth.json` + `config.toml`)
4. Restart or open new terminal to ensure it takes effect
5. To switch back to official login, select "Official Login" from presets and switch; after restarting terminal, follow the official login process
### Basic Usage
### MCP Configuration Guide (v3.5.x)
1. **Add Provider**: Click "Add Provider" → Choose preset or create custom configuration
2. **Switch Provider**:
- Main UI: Select provider → Click "Enable"
- System Tray: Click provider name directly (instant effect)
3. **Takes Effect**: Restart your terminal or Claude Code / Codex / Gemini clients to apply changes
4. **Back to Official**: Select the "Official Login" preset (Claude/Codex) or "Google Official" preset (Gemini), restart the corresponding client, then follow its login/OAuth flow
- Management Location: All MCP server definitions are centrally saved in `~/.cc-switch/config.json` (categorized by client `claude` / `codex`)
- Sync Mechanism:
- Enabled Claude MCP servers are projected to `~/.claude.json` (path may vary with override directory)
- Enabled Codex MCP servers are projected to `~/.codex/config.toml`
- Validation & Normalization: Auto-validate field legality (stdio/http) when adding/importing, and auto-fix/populate keys like `id`
- Import Sources: Support importing from `~/.claude.json` and `~/.codex/config.toml`; existing entries only force `enabled=true`, don't override other fields
### MCP Management
### Check for Updates
- **Location**: Click "MCP" button in top-right corner
- **Add Server**:
- Use built-in templates (mcp-fetch, mcp-filesystem, etc.)
- Support stdio / http / sse transport types
- Configure independent MCP servers for different apps
- **Enable/Disable**: Toggle switches to control which servers sync to live config
- **Sync**: Enabled servers auto-sync to each app's live files
- **Import/Export**: Import existing MCP servers from Claude/Codex/Gemini config files
- Click "Check for Updates" in Settings. If built-in Updater config is available, it will detect and download directly; otherwise, it will fall back to opening the Releases page
### Skills Management (v3.7.0 New)
### Codex Guide (SSOT)
- **Location**: Click "Skills" button in top-right corner
- **Discover Skills**:
- Auto-scan pre-configured GitHub repositories (Anthropic official, ComposioHQ, community, etc.)
- Add custom repositories (supports subdirectory scanning)
- **Install Skills**: Click "Install" to one-click install to `~/.claude/skills/`
- **Uninstall Skills**: Click "Uninstall" to safely remove and clean up state
- **Manage Repositories**: Add/remove custom GitHub repositories
- Config Directory: `~/.codex/`
- Live main config: `auth.json` (required), `config.toml` (can be empty)
- API Key Field: Uses `OPENAI_API_KEY` in `auth.json`
- Switching Behavior (no longer writes "copy files"):
- Provider configs are uniformly saved in `~/.cc-switch/config.json`
- When switching, writes target provider back to live files (`auth.json` + `config.toml`)
- Uses "atomic write + rollback on failure" to avoid half-written state; `config.toml` can be empty
- Import Default: When the app has no providers, creates a default entry from existing live main config and sets it as current
- Official Login: Can switch to preset "Codex Official Login", restart terminal and follow official login process
### Prompts Management (v3.7.0 New)
### Claude Code Guide (SSOT)
- **Location**: Click "Prompts" button in top-right corner
- **Create Presets**:
- Create unlimited system prompt presets
- Use Markdown editor to write prompts (syntax highlighting + real-time preview)
- **Switch Presets**: Select preset → Click "Activate" to apply immediately
- **Sync Mechanism**:
- Claude: `~/.claude/CLAUDE.md`
- Codex: `~/.codex/AGENTS.md`
- Gemini: `~/.gemini/GEMINI.md`
- **Protection Mechanism**: Auto-save current prompt content before switching, preserves manual modifications
- Config Directory: `~/.claude/`
- Live main config: `settings.json` (preferred) or legacy-compatible `claude.json`
- API Key Field: `env.ANTHROPIC_AUTH_TOKEN`
- Switching Behavior (no longer writes "copy files"):
- Provider configs are uniformly saved in `~/.cc-switch/config.json`
- When switching, writes target provider JSON directly to live file (preferring `settings.json`)
- When editing current provider, writes live first successfully, then updates app main config to ensure consistency
- Import Default: When the app has no providers, creates a default entry from existing live main config and sets it as current
- Official Login: Can switch to preset "Claude Official Login", restart terminal and use `/login` to complete login
### Configuration Files
### Migration & Archival
**Claude Code**
#### v3.6 Technical Improvements
- Live config: `~/.claude/settings.json` (or `claude.json`)
- API key field: `env.ANTHROPIC_AUTH_TOKEN` or `env.ANTHROPIC_API_KEY`
- MCP servers: `~/.claude.json``mcpServers`
**Internal Optimizations (User Transparent)**:
**Codex**
- **Removed Legacy Migration Logic**: v3.6 removed v1 config auto-migration and copy file scanning logic
-**Impact**: Improved startup performance, cleaner code
-**Compatibility**: v2 format configs are fully compatible, no action required
- ⚠️ **Note**: Users upgrading from v3.1.0 or earlier should first upgrade to v3.2.x or v3.5.x for one-time migration, then upgrade to v3.6
- Live config: `~/.codex/auth.json` (required) + `config.toml` (optional)
- API key field: `OPENAI_API_KEY` in `auth.json`
- MCP servers: `~/.codex/config.toml``[mcp_servers]` tables
- **Command Parameter Standardization**: Backend unified to use `app` parameter (values: `claude` or `codex`)
-**Impact**: More standardized code, friendlier error messages
-**Compatibility**: Frontend fully adapted, users don't need to care about this change
**Gemini**
#### Startup Failure & Recovery
- Live config: `~/.gemini/.env` (API key) + `~/.gemini/settings.json` (auth mode)
- API key field: `GEMINI_API_KEY` or `GOOGLE_GEMINI_API_KEY` in `.env`
- Environment variables: Support `GOOGLE_GEMINI_BASE_URL`, `GEMINI_MODEL`, etc.
- MCP servers: `~/.gemini/settings.json``mcpServers`
- Tray quick switch: Each provider switch rewrites `~/.gemini/.env`, no need to restart Gemini CLI
- Trigger Conditions: Triggered when `~/.cc-switch/config.json` doesn't exist, is corrupted, or fails to parse.
- User Action: Check JSON syntax according to popup prompt, or restore from backup files.
- Backup Location & Rotation: `~/.cc-switch/backups/backup_YYYYMMDD_HHMMSS.json` (keep up to 10, see `src-tauri/src/services/config.rs`).
- Exit Strategy: To protect data safety, the app will show a popup and force exit when the above errors occur; restart after fixing.
**CC Switch Storage (v3.8.0 New Architecture)**
#### Migration Mechanism (v3.2.0+)
- Database (SSOT): `~/.cc-switch/cc-switch.db` (SQLite, stores providers, MCP, Prompts, Skills)
- Local settings: `~/.cc-switch/settings.json` (device-level settings)
- Backups: `~/.cc-switch/backups/` (auto-rotate, keep 10)
- One-time Migration: First launch of v3.2.0+ will scan old "copy files" and merge into `~/.cc-switch/config.json`
- Claude: `~/.claude/settings-*.json` (excluding `settings.json` / legacy `claude.json`)
- Codex: `~/.codex/auth-*.json` and `config-*.toml` (merged in pairs by name)
- Deduplication & Current Item: Deduplicate by "name (case-insensitive) + API Key"; if current is empty, set live merged item as current
- Archival & Cleanup:
- Archive directory: `~/.cc-switch/archive/<timestamp>/<category>/...`
- Delete original copies after successful archival; keep original files on failure (conservative strategy)
- v1 → v2 Structure Upgrade: Additionally generates `~/.cc-switch/config.v1.backup.<timestamp>.json` for rollback
- Note: After migration, daily switch/edit operations are no longer archived; prepare your own backup solution if long-term auditing is needed
### Cloud Sync Setup
## Architecture Overview (v3.6)
1. Go to Settings → "Custom Configuration Directory"
2. Choose your cloud sync folder (Dropbox, OneDrive, iCloud, etc.)
3. Restart app to apply
4. Repeat on other devices to enable cross-device sync
### Architecture Refactoring Highlights (v3.6)
> **Note**: First launch auto-imports existing Claude/Codex configs as default provider.
**Backend Refactoring (Rust)**: Completed 5-phase refactoring
## Architecture Overview
- **Phase 1**: Unified error handling (`AppError` + i18n error messages)
- **Phase 2**: Command layer split by domain (`commands/{provider,mcp,config,settings,plugin,misc}.rs`)
- **Phase 3**: Introduced integration tests and transaction mechanism (config snapshot + failure rollback)
- **Phase 4**: Extracted Service layer (`services/{provider,mcp,config,speedtest}.rs`)
- **Phase 5**: Concurrency optimization (`RwLock` instead of `Mutex`, scoped guard to avoid deadlock)
### Design Principles
**Frontend Refactoring (React + TypeScript)**: Completed 4-stage refactoring
```
┌─────────────────────────────────────────────────────────────┐
│ Frontend (React + TS) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Components │ │ Hooks │ │ TanStack Query │ │
│ │ (UI) │──│ (Bus. Logic) │──│ (Cache/Sync) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└────────────────────────┬────────────────────────────────────┘
│ Tauri IPC
┌────────────────────────▼────────────────────────────────────┐
│ Backend (Tauri + Rust) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Commands │ │ Services │ │ Models/Config │ │
│ │ (API Layer) │──│ (Bus. Layer) │──│ (Data) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└─────────────────────────────────────────────────────────────┘
```
- **Stage 1**: Established test infrastructure (vitest + MSW + @testing-library/react)
- **Stage 2**: Extracted custom hooks (`useProviderActions`, `useMcpActions`, `useSettings`, `useImportExport`, etc.)
- **Stage 3**: Component splitting and business logic extraction
- **Stage 4**: Code cleanup and formatting unification
**Core Design Patterns**
**Test Coverage**:
- **SSOT** (Single Source of Truth): All data stored in `~/.cc-switch/cc-switch.db` (SQLite)
- **Dual-layer Storage**: SQLite for syncable data, JSON for device-level settings
- **Dual-way Sync**: Write to live files on switch, backfill from live when editing active provider
- **Atomic Writes**: Temp file + rename pattern prevents config corruption
- **Concurrency Safe**: Mutex-protected database connection avoids race conditions
- **Layered Architecture**: Clear separation (Commands → Services → DAO → Database)
- 100% Hooks unit test coverage
- Integration tests covering critical flows (App, SettingsDialog, MCP Panel)
- MSW mocking backend API to ensure test independence
**Key Components**
### Layered Architecture
- **ProviderService**: Provider CRUD, switching, backfill, sorting
- **McpService**: MCP server management, import/export, live file sync
- **ConfigService**: Config import/export, backup rotation
- **SpeedtestService**: API endpoint latency measurement
- **Frontend (Renderer)**
- Tech Stack: TypeScript + React 18 + Vite + TailwindCSS 4
- Data Layer: TanStack React Query unified queries and mutations (`@/lib/query`), Tauri API unified wrapper (`@/lib/api`)
- Business Logic Layer: Custom Hooks (`@/hooks`) carry domain logic, components stay simple
- Event Flow: Listen to backend `provider-switched` events, drive UI refresh and tray state consistency
- Organization: Components split by domain (`providers/settings/mcp/ui`)
**v3.6 Refactoring**
- **Backend (Tauri + Rust)**
- **Commands Layer** (Interface Layer): `src-tauri/src/commands/*` split by domain, only responsible for parameter parsing and permission validation
- **Services Layer** (Business Layer): `src-tauri/src/services/*` carry core logic, reusable and testable
- `ProviderService`: Provider CRUD, switch, backfill, sorting
- `McpService`: MCP server management, import/export, sync
- `ConfigService`: Config file import/export, backup/restore
- `SpeedtestService`: API endpoint latency testing
- **Models & State**:
- `provider.rs`: Domain models (`Provider`, `ProviderManager`, `ProviderMeta`)
- `app_config.rs`: Multi-app config (`MultiAppConfig`, `AppId`, `McpRoot`)
- `store.rs`: Global state (`AppState` + `RwLock<MultiAppConfig>`)
- **Reliability**:
- Unified error type `AppError` (with localized messages)
- Transactional changes (config snapshot + failure rollback)
- Atomic writes (temp file + rename, avoid half-writes)
- Tray menu & events: Rebuild menu after switch and emit `provider-switched` event to frontend
- **Design Points (SSOT + Dual-way Sync)**
- **Single Source of Truth**: Provider configs centrally stored in `~/.cc-switch/config.json`
- **Write on Switch**: Write target provider config to live files (Claude: `settings.json`; Codex: `auth.json` + `config.toml`)
- **Backfill Mechanism**: Immediately read back live files after switch, update SSOT to protect user manual modifications
- **Directory Switch Sync**: Auto-sync current provider to new directory when changing config directories (perfect WSL environment support)
- **Prioritize Live When Editing**: When editing current provider, prioritize loading live config to ensure display of actually effective configuration
- **Compatibility & Changes**
- Command Parameters Unified: Tauri commands only accept `app` (values: `claude` / `codex`)
- Frontend Types Unified: Use `AppId` to express app identifiers (replacing legacy `AppType` export)
- Backend: 5-phase refactoring (error handling → command split → tests → services → concurrency)
- Frontend: 4-stage refactoring (test infra → hooks → components → cleanup)
- Testing: 100% hooks coverage + integration tests (vitest + MSW)
## Development
@@ -346,12 +389,12 @@ cargo test --features test-hooks
**Test Coverage**:
- Hooks unit tests (100% coverage)
- Hooks unit tests (100% coverage)
- `useProviderActions` - Provider operations
- `useMcpActions` - MCP management
- `useSettings` series - Settings management
- `useImportExport` - Import/export
- Integration tests
- Integration tests
- App main application flow
- SettingsDialog complete interaction
- MCP panel functionality
@@ -371,120 +414,36 @@ pnpm test:unit --coverage
## Tech Stack
### Frontend
**Frontend**: React 18 · TypeScript · Vite · TailwindCSS 4 · TanStack Query v5 · react-i18next · react-hook-form · zod · shadcn/ui · @dnd-kit
- **[React 18](https://react.dev/)** - User interface library
- **[TypeScript](https://www.typescriptlang.org/)** - Type-safe JavaScript
- **[Vite](https://vitejs.dev/)** - Lightning fast frontend build tool
- **[TailwindCSS 4](https://tailwindcss.com/)** - Utility-first CSS framework
- **[TanStack Query v5](https://tanstack.com/query/latest)** - Powerful data fetching and caching
- **[react-i18next](https://react.i18next.com/)** - React internationalization framework
- **[react-hook-form](https://react-hook-form.com/)** - High-performance forms library
- **[zod](https://zod.dev/)** - TypeScript-first schema validation
- **[shadcn/ui](https://ui.shadcn.com/)** - Reusable React components
- **[@dnd-kit](https://dndkit.com/)** - Modern drag and drop toolkit
**Backend**: Tauri 2.8 · Rust · serde · tokio · thiserror · tauri-plugin-updater/process/dialog/store/log
### Backend
- **[Tauri 2.8](https://tauri.app/)** - Cross-platform desktop app framework
- tauri-plugin-updater - Auto update
- tauri-plugin-process - Process management
- tauri-plugin-dialog - File dialogs
- tauri-plugin-store - Persistent storage
- tauri-plugin-log - Logging
- **[Rust](https://www.rust-lang.org/)** - Systems programming language
- **[serde](https://serde.rs/)** - Serialization/deserialization framework
- **[tokio](https://tokio.rs/)** - Async runtime
- **[thiserror](https://github.com/dtolnay/thiserror)** - Error handling derive macro
### Testing Tools
- **[vitest](https://vitest.dev/)** - Fast unit testing framework
- **[MSW](https://mswjs.io/)** - API mocking tool
- **[@testing-library/react](https://testing-library.com/react)** - React testing utilities
**Testing**: vitest · MSW · @testing-library/react
## Project Structure
```
├── src/ # Frontend code (React + TypeScript)
│ ├── components/ # React components
│ ├── providers/ # Provider management components
│ │ │ ├── forms/ # Form sub-components (Claude/Codex fields)
│ │ │ ├── ProviderList.tsx
│ │ │ ├── ProviderForm.tsx
│ │ │ ├── AddProviderDialog.tsx
│ │ │ └── EditProviderDialog.tsx
│ │ ├── settings/ # Settings related components
│ │ │ ├── SettingsDialog.tsx
│ │ │ ├── DirectorySettings.tsx
│ │ │ └── ImportExportSection.tsx
│ │ ├── mcp/ # MCP management components
│ │ │ ├── McpPanel.tsx
│ │ │ ├── McpFormModal.tsx
│ │ │ └── McpWizard.tsx
│ │ └── ui/ # shadcn/ui base components
│ ├── hooks/ # Custom Hooks (business logic layer)
│ │ ├── useProviderActions.ts # Provider operations
│ │ ├── useMcpActions.ts # MCP operations
│ │ ├── useSettings.ts # Settings management
│ │ ├── useImportExport.ts # Import/export
│ │ └── useDirectorySettings.ts # Directory config
├── src/ # Frontend (React + TypeScript)
│ ├── components/ # UI components (providers/settings/mcp/ui)
│ ├── hooks/ # Custom hooks (business logic)
│ ├── lib/
│ │ ├── api/ # Tauri API wrapper (type-safe)
│ │ │ ├── providers.ts # Provider API
│ │ │ ├── settings.ts # Settings API
│ │ │ ├── mcp.ts # MCP API
│ │ │ └── usage.ts # Usage query API
│ │ └── query/ # TanStack Query config
│ ├── queries.ts # Query definitions
│ ├── mutations.ts # Mutation definitions
│ └── queryClient.ts
│ ├── i18n/ # Internationalization resources
│ └── locales/
│ ├── zh/ # Chinese translations
── en/ # English translations
├── config/ # Config & presets
├── claudeProviderPresets.ts # Claude provider presets
├── codexProviderPresets.ts # Codex provider presets
└── mcpPresets.ts # MCP server templates
│ ├── utils/ # Utility functions
│ ├── postChangeSync.ts # Config sync utility
│ └── ...
│ └── types/ # TypeScript type definitions
├── src-tauri/ # Backend code (Rust)
│ ├── src/
│ │ ├── commands/ # Tauri command layer (split by domain)
│ │ │ ├── provider.rs # Provider commands
│ │ │ ├── mcp.rs # MCP commands
│ │ │ ├── config.rs # Config query commands
│ │ │ ├── settings.rs # Settings commands
│ │ │ ├── plugin.rs # Plugin commands
│ │ │ ├── import_export.rs # Import/export commands
│ │ │ └── misc.rs # Misc commands
│ │ ├── services/ # Service layer (business logic)
│ │ │ ├── provider.rs # ProviderService
│ │ │ ├── mcp.rs # McpService
│ │ │ ├── config.rs # ConfigService
│ │ │ └── speedtest.rs # SpeedtestService
│ │ ├── app_config.rs # Config data models
│ │ ├── provider.rs # Provider domain models
│ │ ├── store.rs # Global state management
│ │ ├── mcp.rs # MCP sync & validation
│ │ ├── error.rs # Unified error type
│ │ ├── usage_script.rs # Usage script execution
│ │ ├── claude_plugin.rs # Claude plugin management
│ │ └── lib.rs # App entry point
│ ├── capabilities/ # Tauri permission config
│ └── icons/ # App icons
├── tests/ # Frontend tests (v3.6 new)
│ ├── hooks/ # Hooks unit tests
│ ├── components/ # Component integration tests
│ └── setup.ts # Test config
└── assets/ # Static resources
├── screenshots/ # Interface screenshots
└── partners/ # Partner resources
├── logos/ # Partner logos
└── banners/ # Partner banners/promotional images
├── i18n/locales/ # Translations (zh/en)
├── config/ # Presets (providers/mcp)
└── types/ # TypeScript definitions
├── src-tauri/ # Backend (Rust)
│ └── src/
│ ├── commands/ # Tauri command layer (by domain)
── services/ # Business logic layer
├── app_config.rs # Config data models
├── provider.rs # Provider domain models
├── mcp.rs # MCP sync & validation
└── lib.rs # App entry & tray menu
├── tests/ # Frontend tests
│ ├── hooks/ # Unit tests
│ └── components/ # Integration tests
└── assets/ # Screenshots & partner resources
```
## Changelog
@@ -506,7 +465,7 @@ Before submitting PRs, please ensure:
- Pass type check: `pnpm typecheck`
- Pass format check: `pnpm format:check`
- Pass unit tests: `pnpm test:unit`
- Functional PRs should be discussed in the issue area first
- 💡 For new features, please open an issue for discussion before submitting a PR
## Star History
+476
View File
@@ -0,0 +1,476 @@
<div align="center">
# Claude Code / Codex / Gemini CLI オールインワン・アシスタント
[![Version](https://img.shields.io/badge/version-3.8.2-blue.svg)](https://github.com/farion1231/cc-switch/releases)
[![Trending](https://img.shields.io/badge/🔥_TypeScript_Trending-Daily%20%7C%20Weekly%20%7C%20Monthly-ff6b6b.svg)](https://github.com/trending/typescript)
[![Platform](https://img.shields.io/badge/platform-Windows%20%7C%20macOS%20%7C%20Linux-lightgrey.svg)](https://github.com/farion1231/cc-switch/releases)
[![Built with Tauri](https://img.shields.io/badge/built%20with-Tauri%202-orange.svg)](https://tauri.app/)
[![Downloads](https://img.shields.io/endpoint?url=https://api.pinstudios.net/api/badges/downloads/farion1231/cc-switch/total)](https://github.com/farion1231/cc-switch/releases/latest)
<a href="https://trendshift.io/repositories/15372" target="_blank"><img src="https://trendshift.io/api/badge/repositories/15372" alt="farion1231%2Fcc-switch | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
[English](README.md) | [中文](README_ZH.md) | 日本語 | [Changelog](CHANGELOG.md) | [v3.8.0 リリースノート](docs/release-note-v3.8.0-en.md)
**プロバイダスイッチャーから AI CLI 一体型管理プラットフォームへ**
Claude Code・Codex・Gemini CLI のプロバイダ設定、MCP サーバー、Skills 拡張、システムプロンプトを統合管理。
</div>
## ❤️スポンサー
![Zhipu GLM](assets/partners/banners/glm-en.jpg)
本プロジェクトは Z.ai の GLM CODING PLAN による支援を受けています。
GLM CODING PLAN は AI コーディング向けのサブスクリプションで、月額わずか 3 ドルから。Claude Code、Cline、Roo Code など 10 以上の人気 AI コーディングツールでフラッグシップモデル GLM-4.6 を利用でき、速く安定した開発体験を提供します。
[このリンク](https://z.ai/subscribe?ic=8JVLJQFSKB) から申し込むと 10% オフになります!
---
<table>
<tr>
<td width="180"><img src="assets/partners/logos/packycode.png" alt="PackyCode" width="150"></td>
<td>PackyCode のご支援に感謝します!PackyCode は Claude Code、Codex、Gemini などのリレーサービスを提供する信頼性の高い API 中継プラットフォームです。本ソフト利用者向けに特別割引があります:<a href="https://www.packyapi.com/register?aff=cc-switch">このリンク</a>で登録し、チャージ時に「cc-switch」クーポンを入力すると 10% オフになります。</td>
</tr>
<tr>
<td width="180"><img src="assets/partners/logos/sds-en.png" alt="ShanDianShuo" width="150"></td>
<td>ShanDianShuo のご支援に感謝します!ShanDianShuo はローカルファーストの音声入力ツールで、ミリ秒遅延・データは端末から外に出ず・キーボード入力の 4 倍の速度・AI 自動補正・プライバシー優先で完全無料。Claude Code と組み合わせればコーディング効率が倍増します。<a href="https://www.shandianshuo.cn">Mac/Win 版を無料ダウンロード</a></td>
</tr>
</table>
## スクリーンショット
| メイン画面 | プロバイダ追加 |
| :-------------------------------------------: | :----------------------------------------------: |
| ![メイン画面](assets/screenshots/main-ja.png) | ![プロバイダ追加](assets/screenshots/add-ja.png) |
## 特長
### 現在のバージョン:v3.8.2 | [完全な更新履歴](CHANGELOG.md) | [リリースノート](docs/release-note-v3.8.0-en.md)
**v3.8.0 メジャーアップデート (2025-11-28)**
**永続化アーキテクチャ刷新 & 新 UI**
- **SQLite + JSON 二層構造**
- JSON 単独保存から SQLite + JSON の二層構造へ移行
- 同期対象データ(プロバイダ、MCP、Prompts、Skills)は SQLite に保存
- デバイス固有データ(ウィンドウ状態、ローカルパス)は JSON に保存
- 将来のクラウド同期の土台を用意
- DB マイグレーション用にスキーマバージョンを管理
- **新しいユーザーインターフェース**
- レイアウトを全面再設計
- コンポーネントスタイルとアニメーションを統一
- 視覚的な階層を最適化
- ブラウザ互換性向上のため Tailwind CSS を v4 から v3.4 にダウングレード
- **日本語対応**
- UI が中国語/英語/日本語の 3 言語対応に
- **自動起動**
- 設定画面でワンクリック ON/OFF
- プラットフォームネイティブ APIRegistry/LaunchAgent/XDG autostart)を使用
- **Skills 再帰スキャン**
- 多階層ディレクトリをサポート
- リポジトリが異なる同名スキルを許可
- **重要なバグ修正**
- プロバイダ更新時にカスタムエンドポイントが失われる問題を修正
- Gemini 設定の書き込み問題を修正
- Linux WebKitGTK の描画問題を修正
**v3.7.0 ハイライト**
**6 つのコア機能、18,000 行超の新コード**
- **Gemini CLI 統合**
- Claude Code / Codex / Gemini の 3 番目のサポート AI CLI
- 2 つの設定ファイル(`.env` + `settings.json`)に対応
- MCP サーバー管理を完備
- プリセット:Google 公式(OAuth/ PackyCode / カスタム
- **Claude Skills 管理システム**
- GitHub リポジトリを自動スキャン(3 つのキュレーション済みリポジトリを同梱)
- `~/.claude/skills/` へワンクリックでインストール/アンインストール
- カスタムリポジトリ + サブディレクトリスキャンをサポート
- ライフサイクル管理(検出/インストール/更新)を完備
- **Prompts 管理システム**
- 無制限のシステムプロンプトプリセットを作成
- Markdown エディタ(CodeMirror 6 + リアルタイムプレビュー)付き
- スマートなバックフィル保護で手動変更を保持
- 複数アプリに同時対応(Claude: `CLAUDE.md` / Codex: `AGENTS.md` / Gemini: `GEMINI.md`
- **MCP v3.7.0 統合アーキテクチャ**
- 1 つのパネルで 3 アプリの MCP を管理
- 新たに SSEServer-Sent Events)トランスポートを追加
- スマート JSON パーサー + Codex TOML 自動修正
- 双方向のインポート/エクスポート + 双方向同期
- **ディープリンクプロトコル**
- `ccswitch://` を全プラットフォームで登録
- 共有リンクからプロバイダ設定をワンクリックでインポート
- セキュリティ検証 + ライフサイクル統合
- **環境変数の競合検知**
- Claude/Codex/Gemini/MCP 間の設定競合を自動検出
- 競合表示 + 解決ガイド
- 上書き前の警告 + バックアップ
**コア機能**
- **プロバイダ管理**Claude Code、Codex、Gemini の API 設定をワンクリックで切り替え
- **速度テスト**:エンドポイント遅延を計測し、品質を可視化
- **インポート/エクスポート**:設定をバックアップ・復元(最新 10 件を自動ローテーション)
- **多言語対応**:UI/エラー/トレイを含む中国語・英語・日本語ローカライズ
- **Claude プラグイン同期**:Claude プラグイン設定をワンクリックで適用/復元
**v3.6 ハイライト**
- プロバイダの複製とドラッグ&ドロップ並び替え
- 複数エンドポイント管理とカスタム設定ディレクトリ(クラウド同期準備済み)
- 4 階層のモデル設定(Haiku/Sonnet/Opus/Custom
- WSL 環境をサポートし、ディレクトリ変更時に自動同期
- Hooks テスト 100% カバレッジ + アーキテクチャ全面リファクタ
**システム機能**
- クイックスイッチ付きシステムトレイ
- シングルインスタンス常駐
- ビルトイン自動アップデータ
- ロールバック保護付きのアトミック書き込み
## ダウンロード & インストール
### システム要件
- **Windows**: Windows 10 以上
- **macOS**: macOS 10.15 (Catalina) 以上
- **Linux**: Ubuntu 22.04+ / Debian 11+ / Fedora 34+ など主要ディストリビューション
### Windows ユーザー
[Releases](../../releases) ページから最新版の `CC-Switch-v{version}-Windows.msi` インストーラー、またはポータブル版 `CC-Switch-v{version}-Windows-Portable.zip` をダウンロード。
### macOS ユーザー
**方法 1: Homebrew でインストール(推奨)**
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
アップデート:
```bash
brew upgrade --cask cc-switch
```
**方法 2: 手動ダウンロード**
[Releases](../../releases) から `CC-Switch-v{version}-macOS.zip` をダウンロードして展開。
> **注意**: 開発者アカウント未登録のため、初回起動時に「開発元を確認できません」と表示される場合があります。一度閉じてから「システム設定」→「プライバシーとセキュリティ」→「このまま開く」をクリックしてください。以降は通常通り起動できます。
### ArchLinux ユーザー
**paru でインストール(推奨)**
```bash
paru -S cc-switch-bin
```
### Linux ユーザー
[Releases](../../releases) から最新版の `CC-Switch-v{version}-Linux.deb` または `CC-Switch-v{version}-Linux.AppImage` をダウンロード。
## クイックスタート
### 基本的な使い方
1. **プロバイダ追加**:「Add Provider」をクリック → プリセットを選ぶかカスタム設定を作成
2. **プロバイダ切り替え**:
- メイン UI: プロバイダを選択 → 「Enable」をクリック
- システムトレイ: プロバイダ名をクリック(即時反映)
3. **反映**: ターミナルや Claude Code / Codex / Gemini クライアントを再起動して適用
4. **公式設定に戻す**: 「Official Login」プリセット(Claude/Codex)または「Google Official」プリセット(Gemini)を選び、対応クライアントを再起動してログイン/OAuth を実行
### MCP 管理
- **入口**: 右上の「MCP」ボタンをクリック
- **サーバー追加**:
- 組み込みテンプレート(mcp-fetch、mcp-filesystem など)を使用
- stdio / http / sse の各トランスポートをサポート
- アプリごとに独立した MCP を設定可能
- **有効/無効**: トグルでライブ設定への同期を切り替え
- **同期**: 有効なサーバーは各アプリのライブファイルへ自動同期
- **インポート/エクスポート**: Claude/Codex/Gemini の設定ファイルから既存 MCP を取り込み
### Skills 管理 (v3.7.0 新機能)
- **入口**: 右上の「Skills」ボタンをクリック
- **スキル探索**:
- 事前設定済みの GitHub リポジトリを自動スキャン(Anthropic 公式、ComposioHQ、コミュニティなど)
- カスタムリポジトリを追加(サブディレクトリスキャン対応)
- **インストール**: 「Install」を押すだけで `~/.claude/skills/` に配置
- **アンインストール**: 「Uninstall」で安全に削除と状態クリーンアップ
- **リポジトリ管理**: カスタム GitHub リポジトリを追加/削除
### Prompts 管理 (v3.7.0 新機能)
- **入口**: 右上の「Prompts」ボタンをクリック
- **プリセット作成**:
- 無制限のシステムプロンプトプリセットを作成
- Markdown エディタで記述(シンタックスハイライト + リアルタイムプレビュー)
- **プリセット切り替え**: プリセットを選択 → 「Activate」で即適用
- **同期先**:
- Claude: `~/.claude/CLAUDE.md`
- Codex: `~/.codex/AGENTS.md`
- Gemini: `~/.gemini/GEMINI.md`
- **保護機構**: 切り替え前に現在の内容を自動保存し、手動変更を保持
### 設定ファイルパス
**Claude Code**
- ライブ設定: `~/.claude/settings.json`(または `claude.json`
- API キーフィールド: `env.ANTHROPIC_AUTH_TOKEN` または `env.ANTHROPIC_API_KEY`
- MCP サーバー: `~/.claude.json``mcpServers`
**Codex**
- ライブ設定: `~/.codex/auth.json`(必須)+ `config.toml`(任意)
- API キーフィールド: `auth.json` 内の `OPENAI_API_KEY`
- MCP サーバー: `~/.codex/config.toml``[mcp_servers]` テーブル
**Gemini**
- ライブ設定: `~/.gemini/.env`API キー)+ `~/.gemini/settings.json`(認証モード)
- API キーフィールド: `.env` 内の `GEMINI_API_KEY` または `GOOGLE_GEMINI_API_KEY`
- 環境変数: `GOOGLE_GEMINI_BASE_URL``GEMINI_MODEL` などをサポート
- MCP サーバー: `~/.gemini/settings.json``mcpServers`
- トレイでのクイックスイッチ: プロバイダ切り替えごとに `~/.gemini/.env` を書き換えるため Gemini CLI の再起動は不要
**CC Switch 保存先 (v3.8.0 新アーキテクチャ)**
- データベース (SSOT): `~/.cc-switch/cc-switch.db`SQLite。プロバイダ、MCP、Prompts、Skills を保存)
- ローカル設定: `~/.cc-switch/settings.json`(デバイスレベル設定)
- バックアップ: `~/.cc-switch/backups/`(自動ローテーション、最新 10 件を保持)
### クラウド同期の設定
1. 設定 → 「Custom Configuration Directory」へ進む
2. クラウド同期フォルダ(Dropbox、OneDrive、iCloud など)を選択
3. アプリを再起動して反映
4. 他のデバイスでも同じフォルダを指定すればクロスデバイス同期が有効に
> **補足**: 初回起動時に既存の Claude/Codex 設定をデフォルトプロバイダとして自動インポートします。
## アーキテクチャ概要
### 設計原則
```
┌─────────────────────────────────────────────────────────────┐
│ Frontend (React + TS) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Components │ │ Hooks │ │ TanStack Query │ │
│ │ (UI) │──│ (Bus. Logic) │──│ (Cache/Sync) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└────────────────────────┬────────────────────────────────────┘
│ Tauri IPC
┌────────────────────────▼────────────────────────────────────┐
│ Backend (Tauri + Rust) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Commands │ │ Services │ │ Models/Config │ │
│ │ (API Layer) │──│ (Bus. Layer) │──│ (Data) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└─────────────────────────────────────────────────────────────┘
```
**コア設計パターン**
- **SSOT** (Single Source of Truth): すべてのデータを `~/.cc-switch/cc-switch.db`SQLite)に集約
- **二層ストレージ**: 同期データは SQLite、デバイスデータは JSON
- **双方向同期**: 切り替え時はライブファイルへ書き込み、編集時はアクティブプロバイダから逆同期
- **アトミック書き込み**: 一時ファイル + rename パターンで設定破損を防止
- **並行安全**: Mutex で保護された DB 接続でレースを防ぐ
- **レイヤードアーキテクチャ**: Commands → Services → DAO → Database を明確に分離
**主要コンポーネント**
- **ProviderService**: プロバイダの CRUD、切り替え、バックフィル、ソート
- **McpService**: MCP サーバー管理、インポート/エクスポート、ライブファイル同期
- **ConfigService**: 設定のインポート/エクスポート、バックアップローテーション
- **SpeedtestService**: API エンドポイントの遅延計測
**v3.6 リファクタリング**
- バックエンド: エラーハンドリング → コマンド分割 → テスト → サービス層 → 並行性の 5 フェーズ
- フロントエンド: テスト基盤 → hooks → コンポーネント → クリーンアップの 4 ステージ
- テスト: hooks 100% カバレッジ + 統合テスト(vitest + MSW
## 開発
### 開発環境
- Node.js 18+
- pnpm 8+
- Rust 1.85+
- Tauri CLI 2.8+
### 開発コマンド
```bash
# 依存関係をインストール
pnpm install
# ホットリロード付き開発モード
pnpm dev
# 型チェック
pnpm typecheck
# コード整形
pnpm format
# フォーマット検証
pnpm format:check
# フロントエンド単体テスト
pnpm test:unit
# ウォッチモード(開発に推奨)
pnpm test:unit:watch
# アプリをビルド
pnpm build
# デバッグビルド
pnpm tauri build --debug
```
### Rust バックエンド開発
```bash
cd src-tauri
# Rust コード整形
cargo fmt
# clippy チェック
cargo clippy
# バックエンドテスト
cargo test
# 特定テストのみ実行
cargo test test_name
# test-hooks フィーチャー付きでテスト
cargo test --features test-hooks
```
### テストガイド (v3.6)
**フロントエンドテスト**:
- テストフレームワークに **vitest** を使用
- **MSW (Mock Service Worker)** で Tauri API 呼び出しをモック
- コンポーネントテストに **@testing-library/react** を採用
**テストカバレッジ**:
- Hooks の単体テスト(100% カバレッジ)
- `useProviderActions` - プロバイダ操作
- `useMcpActions` - MCP 管理
- `useSettings` 系 - 設定管理
- `useImportExport` - インポート/エクスポート
- 統合テスト
- アプリのメインフロー
- SettingsDialog の一連操作
- MCP パネルの機能
**テスト実行**:
```bash
# 全テストを実行
pnpm test:unit
# ウォッチモード(自動再実行)
pnpm test:unit:watch
# カバレッジレポート付き
pnpm test:unit --coverage
```
## 技術スタック
**フロントエンド**: React 18 · TypeScript · Vite · TailwindCSS 4 · TanStack Query v5 · react-i18next · react-hook-form · zod · shadcn/ui · @dnd-kit
**バックエンド**: Tauri 2.8 · Rust · serde · tokio · thiserror · tauri-plugin-updater/process/dialog/store/log
**テスト**: vitest · MSW · @testing-library/react
## プロジェクト構成
```
├── src/ # フロントエンド (React + TypeScript)
│ ├── components/ # UI コンポーネント (providers/settings/mcp/ui)
│ ├── hooks/ # ビジネスロジック用カスタムフック
│ ├── lib/
│ │ ├── api/ # Tauri API ラッパー (型安全)
│ │ └── query/ # TanStack Query 設定
│ ├── i18n/locales/ # 翻訳 (zh/en)
│ ├── config/ # プリセット (providers/mcp)
│ └── types/ # TypeScript 型定義
├── src-tauri/ # バックエンド (Rust)
│ └── src/
│ ├── commands/ # Tauri コマンド層 (ドメイン別)
│ ├── services/ # ビジネスロジック層
│ ├── app_config.rs # 設定モデル
│ ├── provider.rs # プロバイダドメインモデル
│ ├── mcp.rs # MCP 同期 & 検証
│ └── lib.rs # アプリエントリ & トレイメニュー
├── tests/ # フロントエンドテスト
│ ├── hooks/ # 単体テスト
│ └── components/ # 統合テスト
└── assets/ # スクリーンショット & スポンサーリソース
```
## 更新履歴
詳細は [CHANGELOG.md](CHANGELOG.md) をご覧ください。
## 旧 Electron 版
[Releases](../../releases) に v2.0.3 の Electron 旧版を残しています。
旧版コードが必要な場合は `electron-legacy` ブランチを取得してください。
## 貢献
Issue や提案を歓迎します!
PR を送る前に以下をご確認ください:
- 型チェック: `pnpm typecheck`
- フォーマットチェック: `pnpm format:check`
- 単体テスト: `pnpm test:unit`
- 💡 新機能の場合は、事前に Issue でディスカッションしていただけると助かります
## Star History
[![Star History Chart](https://api.star-history.com/svg?repos=farion1231/cc-switch&type=Date)](https://www.star-history.com/#farion1231/cc-switch&Date)
## ライセンス
MIT © Jason Young
+249 -290
View File
@@ -1,99 +1,151 @@
# Claude Code & Codex 供应商管理器
<div align="center">
[![Version](https://img.shields.io/badge/version-3.6.0-blue.svg)](https://github.com/farion1231/cc-switch/releases)
# Claude Code / Codex / Gemini CLI 全方位辅助工具
[![Version](https://img.shields.io/badge/version-3.8.2-blue.svg)](https://github.com/farion1231/cc-switch/releases)
[![Trending](https://img.shields.io/badge/🔥_TypeScript_Trending-Daily%20%7C%20Weekly%20%7C%20Monthly-ff6b6b.svg)](https://github.com/trending/typescript)
[![Platform](https://img.shields.io/badge/platform-Windows%20%7C%20macOS%20%7C%20Linux-lightgrey.svg)](https://github.com/farion1231/cc-switch/releases)
[![Built with Tauri](https://img.shields.io/badge/built%20with-Tauri%202-orange.svg)](https://tauri.app/)
[![Downloads](https://img.shields.io/endpoint?url=https://api.pinstudios.net/api/badges/downloads/farion1231/cc-switch/total)](https://github.com/farion1231/cc-switch/releases/latest)
[English](README.md) | 中文 | [更新日志](CHANGELOG.md)
<a href="https://trendshift.io/repositories/15372" target="_blank"><img src="https://trendshift.io/api/badge/repositories/15372" alt="farion1231%2Fcc-switch | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
一个用于管理和切换 Claude Code 与 Codex 不同供应商配置、MCP的桌面应用。
[English](README.md) | 中文 | [日本語](README_JA.md) | [更新日志](CHANGELOG.md) | [v3.8.0 发布说明](docs/release-note-v3.8.0-zh.md)
**从供应商切换器到 AI CLI 一体化管理平台**
统一管理 Claude Code、Codex 与 Gemini CLI 的供应商配置、MCP 服务器、Skills 扩展和系统提示词。
</div>
## ❤️ 赞助商
## ❤️赞助商
![智谱 GLM](assets/partners/banners/glm-zh.jpg)
感谢智谱AI的 GLM CODING PLAN 赞助了本项目!
GLM CODING PLAN 是专为AI编码打造的订阅套餐每月最低仅需20元,即可在十余款主流AI编码工具如 Claude Code、Cline 中畅享智谱旗舰模型 GLM-4.6,为开发者提供顶尖、高速、稳定的编码体验。
GLM CODING PLAN 是专为AI编码打造的订阅套餐,每月最低仅需20元,即可在十余款主流AI编码工具如 Claude Code、Cline 中畅享智谱旗舰模型 GLM-4.6,为开发者提供顶尖、高速、稳定的编码体验。
CC Switch 已经预设了智谱GLM,只需要填写 key 即可一键导入编程工具。智谱AI为本软件的用户提供了特别优惠,使用[此链接](https://www.bigmodel.cn/claude-code?ic=RRVJPB5SII)购买可以享受九折优惠。
## 更新记录
---
> **v3.6.0** :新增编辑模式(供应商复制、手动排序)、自定义端点管理、使用量查询等功能,优化配置目录切换体验(WSL 环境完美支持),新增多个供应商预设(DMXAPI、Azure Codex、AnyRouter、AiHubMix、MiniMax),完成全栈架构重构和测试体系建设。
<table>
<tr>
<td width="180"><img src="assets/partners/logos/packycode.png" alt="PackyCode" width="150"></td>
<td>感谢 PackyCode 赞助了本项目!PackyCode 是一家稳定、高效的API中转服务商,提供 Claude Code、Codex、Gemini 等多种中转服务。PackyCode 为本软件的用户提供了特别优惠,使用<a href="https://www.packyapi.com/register?aff=cc-switch">此链接</a>注册并在充值时填写"cc-switch"优惠码,可以享受9折优惠。</td>
</tr>
> v3.5.0 :新增 MCP 管理、配置导入/导出、端点速度测试功能,完善国际化覆盖,新增 Longcat、kat-coder 预设,标准化发布文件命名规范。
<tr>
<td width="180"><img src="assets/partners/logos/sds-zh.png" alt="ShanDianShuo" width="150"></td>
<td>感谢闪电说赞助了本项目!闪电说是本地优先的 AI 语音输入法:毫秒级响应,数据不离设备;打字速度提升 4 倍,AI 智能纠错;绝对隐私安全,完全免费,配合 Claude Code 写代码效率翻倍!支持 Mac/Win 双平台,<a href="https://www.shandianshuo.cn">免费下载</a></td>
</tr>
> v3.4.0 :新增 i18next 国际化、对新模型(qwen-3-max, GLM-4.6, DeepSeek-V3.2-Exp)的支持、Claude 插件、单实例守护、托盘最小化及安装器优化等。
> v3.3.0 VS Code Codex 插件一键配置/移除(默认自动同步)、Codex 通用配置片段与自定义向导增强、WSL 环境支持、跨平台托盘与 UI 优化。(该 VS Code 写入功能已在 v3.4.x 停用)
> v3.2.0 :全新 UI、macOS系统托盘、内置更新器、原子写入与回滚、改进暗色样式、单一事实源(SSOT)与一次性迁移/归档。
> v3.1.0 :新增 Codex 供应商管理与一键切换,支持导入当前 Codex 配置为默认供应商,并在内部配置从 v1 → v2 迁移前自动备份(详见下文“迁移与归档”)。
> v3.0.0 重大更新:从 Electron 完全迁移到 Tauri 2.0,应用体积显著降低、启动性能大幅提升。
## 功能特性(v3.6.0
### 核心功能
- **MCP (Model Context Protocol) 管理**:完整的 MCP 服务器配置管理系统
- 支持 stdio 和 http 服务器类型,并提供命令校验
- 内置常用 MCP 服务器模板(如 mcp-fetch 等)
- 实时启用/禁用 MCP 服务器,原子文件写入防止配置损坏
- **配置导入/导出**:备份和恢复你的供应商配置
- 一键导出所有配置到 JSON 文件
- 导入配置时自动验证并备份,自动轮换备份(保留最近 10 个)
- 导入后自动同步到 live 配置文件,确保立即生效
- **端点速度测试**:测试 API 端点响应时间
- 测量不同供应商端点的延迟,可视化连接质量指示器
- 帮助用户选择最快的供应商
- **国际化与语言切换**:完整的 i18next 国际化覆盖(包含错误消息、托盘菜单、所有 UI 组件)
- **Claude 插件同步**:内置按钮可一键应用或恢复 Claude 插件配置,切换供应商后立即生效。
### v3.6 新增功能
- **供应商复制功能**:快速复制现有供应商配置,轻松创建变体配置
- **手动排序功能**:通过拖拽来对供应商进行手动排序
- **自定义端点管理**:支持聚合类供应商的多端点配置
- **使用量查询功能**
- 自动刷新间隔:支持定时自动查询使用量
- 测试脚本 API:测试 JavaScript 脚本是否正确
- 模板系统扩展:自定义空白模板、支持 access token 和 user ID 参数
- **配置编辑器改进**
- 新增 JSON 格式化按钮
- 实时 TOML 语法验证(Codex 配置)
- **配置目录切换自动同步**:切换 Claude/Codex 配置目录(如切换到 WSL 环境)时,自动同步当前供应商到新目录,避免冲突导致配置文件混乱
- **编辑当前供应商时加载 live 配置**:编辑正在使用的供应商时,优先显示实际生效的配置,保护用户手动修改
- **新增供应商预设**DMXAPI、Azure Codex、AnyRouter、AiHubMix、MiniMax
- **合作伙伴推广机制**:支持生态合作伙伴推广(如智谱 GLM Z.ai)
### v3.6 架构改进
- **后端重构**:完成 5 阶段重构(统一错误处理 → 命令层拆分 → 集成测试 → Service 层提取 → 并发优化)
- **前端重构**:完成 4 阶段重构(测试基础设施 → Hooks 提取 → 组件拆分 → 代码清理)
- **测试体系**Hooks 单元测试 100% 覆盖,集成测试覆盖关键流程(vitest + MSW + @testing-library/react
### 系统功能
- **系统托盘与窗口行为**:窗口关闭可最小化到托盘,macOS 支持托盘模式下隐藏/显示 Dock,托盘切换时同步 Claude/Codex/插件状态。
- **单实例**:保证同一时间仅运行一个实例,避免多开冲突。
- **标准化发布命名**:所有平台发布文件使用一致的版本标签命名(macOS: `.tar.gz` / `.zip`Windows: `.msi` / `-Portable.zip`Linux: `.AppImage` / `.deb`)。
</table>
## 界面预览
### 主界面
| 主界面 | 添加供应商 |
| :---------------------------------------: | :------------------------------------------: |
| ![主界面](assets/screenshots/main-zh.png) | ![添加供应商](assets/screenshots/add-zh.png) |
![主界面](assets/screenshots/main-zh.png)
## 功能特性
### 添加供应商
### 当前版本:v3.8.2 | [完整更新日志](CHANGELOG.md)
![添加供应商](assets/screenshots/add-zh.png)
**v3.8.0 重大更新(2025-11-28**
**持久化架构升级 & 全新用户界面**
- **SQLite + JSON 双层架构**
- 从 JSON 文件存储迁移到 SQLite + JSON 双层结构
- 可同步数据(供应商、MCP、Prompts、Skills)存入 SQLite
- 设备级数据(窗口状态、本地路径)保留在 JSON
- 为未来云同步功能奠定基础
- Schema 版本管理支持数据库迁移
- **全新用户界面**
- 完全重新设计的界面布局
- 统一的组件样式和更流畅的动画
- 优化的视觉层次
- Tailwind CSS 从 v4 降级到 v3.4 以提升浏览器兼容性
- **日语支持**
- 新增日语界面支持(现支持中文/英文/日语)
- **开机自启**
- 在设置中一键开启/关闭
- 使用平台原生 API(注册表/LaunchAgent/XDG autostart
- **Skills 递归扫描**
- 支持多层目录结构
- 允许不同仓库的同名技能
- **关键 Bug 修复**
- 修复更新供应商时自定义端点丢失问题
- 修复 Gemini 配置写入问题
- 修复 Linux WebKitGTK 渲染问题
**v3.7.0 亮点**
**六大核心功能,18,000+ 行新增代码**
- **Gemini CLI 集成**
- 第三个支持的 AI CLIClaude Code / Codex / Gemini
- 双文件配置支持(`.env` + `settings.json`
- 完整 MCP 服务器管理
- 预设:Google Official (OAuth) / PackyCode / 自定义
- **Claude Skills 管理系统**
- 从 GitHub 仓库自动扫描技能(预配置 3 个精选仓库)
- 一键安装/卸载到 `~/.claude/skills/`
- 自定义仓库支持 + 子目录扫描
- 完整生命周期管理(发现/安装/更新)
- **Prompts 管理系统**
- 多预设系统提示词管理(无限数量,快速切换)
- 跨应用支持(Claude: `CLAUDE.md` / Codex: `AGENTS.md` / Gemini: `GEMINI.md`
- Markdown 编辑器(CodeMirror 6 + 实时预览)
- 智能回填保护,保留手动修改
- **MCP v3.7.0 统一架构**
- 单一面板管理三个应用的 MCP 服务器
- 新增 SSE (Server-Sent Events) 传输类型
- 智能 JSON 解析器 + Codex TOML 格式自动修正
- 统一导入/导出 + 双向同步
- **深度链接协议**
- `ccswitch://` 协议注册(全平台)
- 通过共享链接一键导入供应商配置
- 安全验证 + 生命周期集成
- **环境变量冲突检测**
- 自动检测跨应用配置冲突(Claude/Codex/Gemini/MCP
- 可视化冲突指示器 + 解决建议
- 覆盖警告 + 更改前备份
**核心功能**
- **供应商管理**:一键切换 Claude Code、Codex 与 Gemini 的 API 配置
- **速度测试**:测量 API 端点延迟,可视化连接质量指示器
- **导入导出**:备份和恢复配置,自动轮换(保留最近 10 个)
- **国际化支持**:完整的中英文本地化(UI、错误、托盘)
- **Claude 插件同步**:一键应用或恢复 Claude 插件配置
**v3.6 亮点**
- 供应商复制 & 拖拽排序
- 多端点管理 & 自定义配置目录(支持云同步)
- 细粒度模型配置(四层:Haiku/Sonnet/Opus/自定义)
- WSL 环境支持,配置目录切换自动同步
- 100% hooks 测试覆盖 & 完整架构重构
**系统功能**
- 系统托盘快速切换
- 单实例守护
- 内置自动更新器
- 原子写入与回滚保护
## 下载安装
@@ -128,152 +180,143 @@ brew upgrade --cask cc-switch
> **注意**:由于作者没有苹果开发者账号,首次打开可能出现"未知开发者"警告,请先关闭,然后前往"系统设置" → "隐私与安全性" → 点击"仍要打开",之后便可以正常打开
### ArchLinux 用户
**通过 paru 安装(推荐)**
```bash
paru -S cc-switch-bin
```
### Linux 用户
从 [Releases](../../releases) 页面下载最新版本的 `CC-Switch-v{版本号}-Linux.deb` 包或者 `CC-Switch-v{版本号}-Linux.AppImage` 安装包。
## 使用说明
## 快速开始
1. 点击"添加供应商"添加你的 API 配置
2. 切换方式:
- 在主界面选择供应商后点击切换
- 或通过“系统托盘(菜单栏)”直接选择目标供应商,立即生效
3. 切换会写入对应应用的“live 配置文件”(Claude:`settings.json`Codex`auth.json` + `config.toml`
4. 重启或新开终端以确保生效
5. 若需切回官方登录,在预设中选择“官方登录”并切换即可;重启终端后按官方流程登录
### 基本使用
### MCP 配置说明(v3.5.x
1. **添加供应商**:点击"添加供应商" → 选择预设或创建自定义配置
2. **切换供应商**
- 主界面:选择供应商 → 点击"启用"
- 系统托盘:直接点击供应商名称(立即生效)
3. **生效方式**:重启终端或 Claude Code / Codex / Gemini 客户端以应用更改
4. **恢复官方登录**:选择"官方登录"预设(Claude/Codex)或"Google 官方"预设(Gemini),重启对应客户端后按照其登录/OAuth 流程操作
- 管理位置:所有 MCP 服务器定义集中保存在 `~/.cc-switch/config.json`(按客户端 `claude` / `codex` 分类)
- 同步机制:
- 启用的 Claude MCP 会投影到 `~/.claude.json`(路径可随覆盖目录而变化)
- 启用的 Codex MCP 会投影到 `~/.codex/config.toml`
- 校验与归一化:新增/导入时自动校验字段合法性(stdio/http),并自动修复/填充 `id` 等键名
- 导入来源:支持从 `~/.claude.json``~/.codex/config.toml` 导入;已存在条目只强制 `enabled=true`,不覆盖其他字段
### MCP 管理
### 检查更新
- **位置**:点击右上角"MCP"按钮
- **添加服务器**
- 使用内置模板(mcp-fetch、mcp-filesystem 等)
- 支持 stdio / http / sse 三种传输类型
- 为不同应用配置独立的 MCP 服务器
- **启用/禁用**:切换开关以控制哪些服务器同步到 live 配置
- **同步**:启用的服务器自动同步到各应用的 live 文件
- **导入/导出**:支持从 Claude/Codex/Gemini 配置文件导入现有 MCP 服务器
- 在“设置”中点击“检查更新”,若内置 Updater 配置可用将直接检测与下载;否则会回退打开 Releases 页面
### Skills 管理(v3.7.0 新增)
### Codex 说明(SSOT
- **位置**:点击右上角"Skills"按钮
- **发现技能**
- 自动扫描预配置的 GitHub 仓库(Anthropic 官方、ComposioHQ、社区等)
- 添加自定义仓库(支持子目录扫描)
- **安装技能**:点击"安装"一键安装到 `~/.claude/skills/`
- **卸载技能**:点击"卸载"安全移除并清理状态
- **管理仓库**:添加/删除自定义 GitHub 仓库
- 配置目录:`~/.codex/`
- live 主配置:`auth.json`(必需)、`config.toml`(可为空)
- API Key 字段:`auth.json` 中使用 `OPENAI_API_KEY`
- 切换行为(不再写“副本文件”):
- 供应商配置统一保存在 `~/.cc-switch/config.json`
- 切换时将目标供应商写回 live 文件(`auth.json` + `config.toml`
- 采用“原子写入 + 失败回滚”,避免半写状态;`config.toml` 可为空
- 导入默认:当该应用无任何供应商时,从现有 live 主配置创建一条默认项并设为当前
- 官方登录:可切换到预设“Codex 官方登录”,重启终端后按官方流程登录
### Prompts 管理(v3.7.0 新增)
### Claude Code 说明(SSOT
- **位置**:点击右上角"Prompts"按钮
- **创建预设**
- 创建无限数量的系统提示词预设
- 使用 Markdown 编辑器编写提示词(语法高亮 + 实时预览)
- **切换预设**:选择预设 → 点击"激活"立即应用
- **同步机制**
- Claude: `~/.claude/CLAUDE.md`
- Codex: `~/.codex/AGENTS.md`
- Gemini: `~/.gemini/GEMINI.md`
- **保护机制**:切换前自动保存当前提示词内容,保留手动修改
- 配置目录:`~/.claude/`
- live 主配置:`settings.json`(优先)或历史兼容 `claude.json`
- API Key 字段:`env.ANTHROPIC_AUTH_TOKEN`
- 切换行为(不再写“副本文件”):
- 供应商配置统一保存在 `~/.cc-switch/config.json`
- 切换时将目标供应商 JSON 直接写入 live 文件(优先 `settings.json`
- 编辑当前供应商时,先写 live 成功,再更新应用主配置,保证一致性
- 导入默认:当该应用无任何供应商时,从现有 live 主配置创建一条默认项并设为当前
- 官方登录:可切换到预设“Claude 官方登录”,重启终端后可使用 `/login` 完成登录
### 配置文件
### 迁移与归档
**Claude Code**
#### v3.6 技术改进
- Live 配置:`~/.claude/settings.json`(或 `claude.json`
- API key 字段:`env.ANTHROPIC_AUTH_TOKEN``env.ANTHROPIC_API_KEY`
- MCP 服务器:`~/.claude.json``mcpServers`
**内部优化(用户无感知)**
**Codex**
- **移除遗留迁移逻辑**:v3.6 移除了 v1 配置自动迁移和副本文件扫描逻辑
-**影响**:启动性能提升,代码更简洁
-**兼容性**:v2 格式配置完全兼容,无需任何操作
- ⚠️ **注意**:从 v3.1.0 或更早版本升级的用户,请先升级到 v3.2.x 或 v3.5.x 完成一次性迁移,再升级到 v3.6
- Live 配置:`~/.codex/auth.json`(必需)+ `config.toml`(可选)
- API key 字段:`auth.json` 中的 `OPENAI_API_KEY`
- MCP 服务器:`~/.codex/config.toml``[mcp_servers]`
- **命令参数标准化**:后端统一使用 `app` 参数(取值:`claude``codex`
-**影响**:代码更规范,错误提示更友好
-**兼容性**:前端已完全适配,用户无需关心此变更
**Gemini**
#### 启动失败与恢复
- Live 配置:`~/.gemini/.env`API Key+ `~/.gemini/settings.json`(保存认证模式)
- API key 字段:`.env` 文件中的 `GEMINI_API_KEY``GOOGLE_GEMINI_API_KEY`
- 环境变量:支持 `GOOGLE_GEMINI_BASE_URL``GEMINI_MODEL` 等自定义变量
- MCP 服务器:`~/.gemini/settings.json``mcpServers`
- 托盘快速切换:每次切换供应商都会重写 `~/.gemini/.env`,无需重启 Gemini CLI 即可生效
- 触发条件:`~/.cc-switch/config.json` 不存在、损坏或解析失败时触发。
- 用户动作:根据弹窗提示检查 JSON 语法,或从备份文件恢复。
- 备份位置与轮换:`~/.cc-switch/backups/backup_YYYYMMDD_HHMMSS.json`(最多保留 10 个,参见 `src-tauri/src/services/config.rs`)。
- 退出策略:为保护数据安全,出现上述错误时应用会弹窗提示并强制退出;修复后重新启动即可。
**CC Switch 存储(v3.8.0 新架构)**
#### v3.2.0 起的迁移机制
- 数据库(SSOT):`~/.cc-switch/cc-switch.db`SQLite,存储供应商、MCP、Prompts、Skills
- 本地设置:`~/.cc-switch/settings.json`(设备级设置)
- 备份:`~/.cc-switch/backups/`(自动轮换,保留 10 个)
- 一次性迁移:首次启动 3.2.0 及以上版本会扫描旧的"副本文件"并合并到 `~/.cc-switch/config.json`
- Claude`~/.claude/settings-*.json`(排除 `settings.json` / 历史 `claude.json`
- Codex`~/.codex/auth-*.json``config-*.toml`(按名称成对合并)
- 去重与当前项:按"名称(忽略大小写)+ API Key"去重;若当前为空,将 live 合并项设为当前
- 归档与清理:
- 归档目录:`~/.cc-switch/archive/<timestamp>/<category>/...`
- 归档成功后删除原副本;失败则保留原文件(保守策略)
- v1 → v2 结构升级:会额外生成 `~/.cc-switch/config.v1.backup.<timestamp>.json` 以便回滚
- 注意:迁移后不再持续归档日常切换/编辑操作,如需长期审计请自备备份方案
### 云同步设置
## 架构总览(v3.6
1. 前往设置 → "自定义配置目录"
2. 选择您的云同步文件夹(Dropbox、OneDrive、iCloud、坚果云等)
3. 重启应用以应用
4. 在其他设备上重复操作以启用跨设备同步
### 架构重构亮点(v3.6
> **注意**:首次启动会自动导入现有 Claude/Codex 配置作为默认供应商。
**后端重构(Rust**:完成 5 阶段重构
## 架构总览
- **Phase 1**:统一错误处理(`AppError` + 国际化错误消息)
- **Phase 2**:命令层按领域拆分(`commands/{provider,mcp,config,settings,plugin,misc}.rs`
- **Phase 3**:引入集成测试和事务机制(配置快照 + 失败回滚)
- **Phase 4**:提取 Service 层(`services/{provider,mcp,config,speedtest}.rs`
- **Phase 5**:并发优化(`RwLock` 替代 `Mutex`,作用域 guard 避免死锁)
### 设计原则
**前端重构(React + TypeScript**:完成 4 阶段重构
```
┌─────────────────────────────────────────────────────────────┐
│ 前端 (React + TS) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Components │ │ Hooks │ │ TanStack Query │ │
│ │ (UI) │──│ (业务逻辑) │──│ (缓存/同步) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└────────────────────────┬────────────────────────────────────┘
│ Tauri IPC
┌────────────────────────▼────────────────────────────────────┐
│ 后端 (Tauri + Rust) │
│ ┌─────────────┐ ┌──────────────┐ ┌──────────────────┐ │
│ │ Commands │ │ Services │ │ Models/Config │ │
│ │ (API 层) │──│ (业务层) │──│ (数据) │ │
│ └─────────────┘ └──────────────┘ └──────────────────┘ │
└─────────────────────────────────────────────────────────────┘
```
- **Stage 1**:建立测试基础设施(vitest + MSW + @testing-library/react
- **Stage 2**:提取自定义 hooks`useProviderActions`, `useMcpActions`, `useSettings`, `useImportExport` 等)
- **Stage 3**:组件拆分和业务逻辑提取
- **Stage 4**:代码清理和格式化统一
**核心设计模式**
**测试覆盖**
- **SSOT**(单一事实源):所有数据存储在 `~/.cc-switch/cc-switch.db`SQLite
- **双层存储**SQLite 存储可同步数据,JSON 存储设备级设置
- **双向同步**:切换时写入 live 文件,编辑当前供应商时从 live 回填
- **原子写入**:临时文件 + 重命名模式防止配置损坏
- **并发安全**:Mutex 保护的数据库连接避免竞态条件
- **分层架构**:清晰分离(Commands → Services → DAO → Database
- Hooks 单元测试 100% 覆盖
- 集成测试覆盖关键流程(App、SettingsDialog、MCP 面板)
- MSW 模拟后端 API,确保测试独立性
**核心组件**
### 分层架构
- **ProviderService**:供应商增删改查、切换、回填、排序
- **McpService**MCP 服务器管理、导入导出、live 文件同步
- **ConfigService**:配置导入导出、备份轮换
- **SpeedtestService**API 端点延迟测量
- **前端(Renderer**
- 技术栈:TypeScript + React 18 + Vite + TailwindCSS 4
- 数据层:TanStack React Query 统一查询与变更(`@/lib/query`),Tauri API 统一封装(`@/lib/api`
- 业务逻辑层:自定义 Hooks`@/hooks`)承载领域逻辑,组件保持简洁
- 事件流:监听后端 `provider-switched` 事件,驱动 UI 刷新与托盘状态一致
- 组织结构:按领域拆分组件(`providers/settings/mcp/ui`
**v3.6 重构**
- **后端(Tauri + Rust**
- **Commands 层**(接口层):`src-tauri/src/commands/*` 按领域拆分,仅负责参数解析和权限校验
- **Services 层**(业务层):`src-tauri/src/services/*` 承载核心逻辑,可复用和测试
- `ProviderService`:供应商增删改查、切换、回填、排序
- `McpService`MCP 服务器管理、导入导出、同步
- `ConfigService`:配置文件导入导出、备份恢复
- `SpeedtestService`API 端点延迟测试
- **模型与状态**
- `provider.rs`:领域模型(`Provider`, `ProviderManager`, `ProviderMeta`
- `app_config.rs`:多应用配置(`MultiAppConfig`, `AppId`, `McpRoot`
- `store.rs`:全局状态(`AppState` + `RwLock<MultiAppConfig>`
- **可靠性**
- 统一错误类型 `AppError`(包含本地化消息)
- 事务式变更(配置快照 + 失败回滚)
- 原子写入(临时文件 + 重命名,避免半写入)
- 托盘菜单与事件:切换后重建菜单并向前端发射 `provider-switched` 事件
- **设计要点(SSOT + 双向同步)**
- **单一事实源**:供应商配置集中存放于 `~/.cc-switch/config.json`
- **切换时写入**:将目标供应商配置写入 live 文件(Claude: `settings.json`Codex: `auth.json` + `config.toml`
- **回填机制**:切换后立即读回 live 文件,更新 SSOT,保护用户手动修改
- **目录切换同步**:修改配置目录时自动同步当前供应商到新目录(WSL 环境完美支持)
- **编辑时优先 live**:编辑当前供应商时,优先加载 live 配置,确保显示实际生效的配置
- **兼容性与变更**
- 命令参数统一:Tauri 命令仅接受 `app`(值为 `claude` / `codex`
- 前端类型统一:使用 `AppId` 表达应用标识(替代历史 `AppType` 导出)
- 后端:5 阶段重构(错误处理 → 命令拆分 → 测试 → 服务 → 并发)
- 前端:4 阶段重构(测试基础 → hooks → 组件 → 清理)
- 测试:100% hooks 覆盖 + 集成测试(vitest + MSW
## 开发
@@ -346,12 +389,12 @@ cargo test --features test-hooks
**测试覆盖**
- Hooks 单元测试(100% 覆盖)
- Hooks 单元测试(100% 覆盖)
- `useProviderActions` - 供应商操作
- `useMcpActions` - MCP 管理
- `useSettings` 系列 - 设置管理
- `useImportExport` - 导入导出
- 集成测试
- 集成测试
- App 主应用流程
- SettingsDialog 完整交互
- MCP 面板功能
@@ -371,120 +414,36 @@ pnpm test:unit --coverage
## 技术栈
### 前端
**前端**React 18 · TypeScript · Vite · TailwindCSS 4 · TanStack Query v5 · react-i18next · react-hook-form · zod · shadcn/ui · @dnd-kit
- **[React 18](https://react.dev/)** - 用户界面库
- **[TypeScript](https://www.typescriptlang.org/)** - 类型安全的 JavaScript
- **[Vite](https://vitejs.dev/)** - 极速的前端构建工具
- **[TailwindCSS 4](https://tailwindcss.com/)** - 实用优先的 CSS 框架
- **[TanStack Query v5](https://tanstack.com/query/latest)** - 强大的数据获取与缓存
- **[react-i18next](https://react.i18next.com/)** - React 国际化框架
- **[react-hook-form](https://react-hook-form.com/)** - 高性能表单库
- **[zod](https://zod.dev/)** - TypeScript 优先的模式验证
- **[shadcn/ui](https://ui.shadcn.com/)** - 可复用的 React 组件
- **[@dnd-kit](https://dndkit.com/)** - 现代拖拽工具包
**后端**Tauri 2.8 · Rust · serde · tokio · thiserror · tauri-plugin-updater/process/dialog/store/log
### 后端
- **[Tauri 2.8](https://tauri.app/)** - 跨平台桌面应用框架
- tauri-plugin-updater - 自动更新
- tauri-plugin-process - 进程管理
- tauri-plugin-dialog - 文件对话框
- tauri-plugin-store - 持久化存储
- tauri-plugin-log - 日志记录
- **[Rust](https://www.rust-lang.org/)** - 系统级编程语言
- **[serde](https://serde.rs/)** - 序列化/反序列化框架
- **[tokio](https://tokio.rs/)** - 异步运行时
- **[thiserror](https://github.com/dtolnay/thiserror)** - 错误处理派生宏
### 测试工具
- **[vitest](https://vitest.dev/)** - 快速的单元测试框架
- **[MSW](https://mswjs.io/)** - API mock 工具
- **[@testing-library/react](https://testing-library.com/react)** - React 测试工具
**测试**vitest · MSW · @testing-library/react
## 项目结构
```
├── src/ # 前端代码 (React + TypeScript)
│ ├── components/ # React 组件
│ ├── providers/ # 供应商管理组件
│ │ │ ├── forms/ # 表单子组件(Claude/Codex 字段)
│ │ │ ├── ProviderList.tsx
│ │ │ ├── ProviderForm.tsx
│ │ │ ├── AddProviderDialog.tsx
│ │ │ └── EditProviderDialog.tsx
│ │ ├── settings/ # 设置相关组件
│ │ │ ├── SettingsDialog.tsx
│ │ │ ├── DirectorySettings.tsx
│ │ │ └── ImportExportSection.tsx
│ │ ├── mcp/ # MCP 管理组件
│ │ │ ├── McpPanel.tsx
│ │ │ ├── McpFormModal.tsx
│ │ │ └── McpWizard.tsx
│ │ └── ui/ # shadcn/ui 基础组件
│ ├── hooks/ # 自定义 Hooks(业务逻辑层)
│ │ ├── useProviderActions.ts # 供应商操作
│ │ ├── useMcpActions.ts # MCP 操作
│ │ ├── useSettings.ts # 设置管理
│ │ ├── useImportExport.ts # 导入导出
│ │ └── useDirectorySettings.ts # 目录配置
├── src/ # 前端 (React + TypeScript)
│ ├── components/ # UI 组件 (providers/settings/mcp/ui)
│ ├── hooks/ # 自定义 hooks (业务逻辑)
│ ├── lib/
│ │ ├── api/ # Tauri API 封装(类型安全)
│ │ │ ├── providers.ts # 供应商 API
│ │ │ ├── settings.ts # 设置 API
│ │ │ ├── mcp.ts # MCP API
│ │ │ └── usage.ts # 用量查询 API
│ │ └── query/ # TanStack Query 配置
│ ├── queries.ts # 查询定义
│ ├── mutations.ts # 变更定义
│ │ └── queryClient.ts
│ ├── i18n/ # 国际化资源
│ │ └── locales/
│ │ ├── zh/ # 中文翻译
│ │ └── en/ # 英文翻译
│ ├── config/ # 配置与预设
│ │ ├── claudeProviderPresets.ts # Claude 供应商预设
│ │ ├── codexProviderPresets.ts # Codex 供应商预设
│ │ └── mcpPresets.ts # MCP 服务器模板
│ ├── utils/ # 工具函数
│ │ ├── postChangeSync.ts # 配置同步工具
│ │ └── ...
├── i18n/locales/ # 翻译 (zh/en)
├── config/ # 预设 (providers/mcp)
│ └── types/ # TypeScript 类型定义
├── src-tauri/ # 后端代码 (Rust)
── src/
├── commands/ # Tauri 命令层(按领域拆分
│ │ ├── provider.rs # 供应商命令
│ │ ├── mcp.rs # MCP 命令
│ │ ├── config.rs # 配置查询命令
│ │ ├── settings.rs # 设置命令
│ ├── plugin.rs # 插件命令
│ │ │ ├── import_export.rs # 导入导出命令
└── misc.rs # 杂项命令
│ ├── services/ # Service 层(业务逻辑)
│ │ │ ├── provider.rs # ProviderService
│ │ │ ├── mcp.rs # McpService
│ │ │ ├── config.rs # ConfigService
│ │ │ └── speedtest.rs # SpeedtestService
│ │ ├── app_config.rs # 配置数据模型
│ │ ├── provider.rs # 供应商领域模型
│ │ ├── store.rs # 全局状态管理
│ │ ├── mcp.rs # MCP 同步与校验
│ │ ├── error.rs # 统一错误类型
│ │ ├── usage_script.rs # 用量脚本执行
│ │ ├── claude_plugin.rs # Claude 插件管理
│ │ └── lib.rs # 应用入口
│ ├── capabilities/ # Tauri 权限配置
│ └── icons/ # 应用图标
├── tests/ # 前端测试(v3.6 新增)
│ ├── hooks/ # Hooks 单元测试
│ ├── components/ # 组件集成测试
│ └── setup.ts # 测试配置
└── assets/ # 静态资源
├── screenshots/ # 界面截图
└── partners/ # 合作商资源
├── logos/ # 合作商 Logo
└── banners/ # 合作商横幅/宣传图
├── src-tauri/ # 后端 (Rust)
── src/
├── commands/ # Tauri 命令层(按领域)
├── services/ # 业务逻辑层
├── app_config.rs # 配置数据模型
├── provider.rs # 供应商领域模型
├── mcp.rs # MCP 同步与校验
└── lib.rs # 应用入口 & 托盘菜单
├── tests/ # 前端测试
├── hooks/ # 单元测试
└── components/ # 集成测试
└── assets/ # 截图 & 合作商资源
```
## 更新日志
@@ -506,7 +465,7 @@ pnpm test:unit --coverage
- 通过类型检查:`pnpm typecheck`
- 通过格式检查:`pnpm format:check`
- 通过单元测试:`pnpm test:unit`
- 功能性 PR 请先经过 issue 讨论
- 💡 新功能开发前,欢迎先开 issue 讨论实现方案
## Star History
-76
View File
@@ -1,76 +0,0 @@
# CC Switch 国际化功能说明
## 已完成的工作
1. **安装依赖**:添加了 `react-i18next``i18next`
2. **配置国际化**:在 `src/i18n/` 目录下创建了配置文件
3. **翻译文件**:创建了英文和中文翻译文件
4. **组件更新**:替换了主要组件中的硬编码文案
5. **语言切换器**:添加了语言切换按钮
## 文件结构
```
src/
├── i18n/
│ ├── index.ts # 国际化配置文件
│ └── locales/
│ ├── en.json # 英文翻译
│ └── zh.json # 中文翻译
├── components/
│ └── LanguageSwitcher.tsx # 语言切换组件
└── main.tsx # 导入国际化配置
```
## 默认语言设置
- **默认语言**:英文 (en)
- **回退语言**:英文 (en)
## 使用方式
1. 在组件中导入 `useTranslation`
```tsx
import { useTranslation } from 'react-i18next';
function MyComponent() {
const { t } = useTranslation();
return <div>{t('common.save')}</div>;
}
```
2. 切换语言:
```tsx
const { i18n } = useTranslation();
i18n.changeLanguage('zh'); // 切换到中文
```
## 翻译键结构
- `common.*` - 通用文案(保存、取消、设置等)
- `header.*` - 头部相关文案
- `provider.*` - 供应商相关文案
- `notifications.*` - 通知消息
- `settings.*` - 设置页面文案
- `apps.*` - 应用名称
- `console.*` - 控制台日志信息
## 测试功能
应用已添加了语言切换按钮(地球图标),点击可以在中英文之间切换,验证国际化功能是否正常工作。
## 已更新的组件
- ✅ App.tsx - 主应用组件
- ✅ ConfirmDialog.tsx - 确认对话框
- ✅ AddProviderModal.tsx - 添加供应商弹窗
- ✅ EditProviderModal.tsx - 编辑供应商弹窗
- ✅ ProviderList.tsx - 供应商列表
- ✅ LanguageSwitcher.tsx - 语言切换器
- ✅ settings/SettingsDialog.tsx - 设置对话框
## 注意事项
1. 所有新的文案都应该添加到翻译文件中,而不是硬编码
2. 翻译键名应该有意义且结构化
3. 可以通过修改 `src/i18n/index.ts` 中的 `lng` 配置来更改默认语言
Binary file not shown.

After

Width:  |  Height:  |  Size: 8.1 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 179 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 6.5 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 185 KiB

After

Width:  |  Height:  |  Size: 206 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 216 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 203 KiB

After

Width:  |  Height:  |  Size: 212 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 204 KiB

After

Width:  |  Height:  |  Size: 227 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 226 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 205 KiB

After

Width:  |  Height:  |  Size: 225 KiB

+2715
View File
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
+249
View File
@@ -0,0 +1,249 @@
## Major architecture refactoring with enhanced config sync and data protection
**[中文更新说明 Chinese Documentation →](https://github.com/farion1231/cc-switch/blob/main/docs/release-note-v3.6.0-zh.md)**
---
## What's New
### Edit Mode & Provider Management
- **Provider Duplication** - Quickly duplicate existing provider configurations to create variants with one click
- **Manual Sorting** - Drag and drop to reorder providers, with visual push effect animations. Thanks to @ZyphrZero
- **Edit Mode Toggle** - Show/hide drag handles to optimize editing experience
### Custom Endpoint Management
- **Multi-Endpoint Configuration** - Support for aggregator providers with multiple API endpoints
- **Endpoint Input Visibility** - Shows endpoint field for all non-official providers automatically
### Usage Query Enhancements
- **Auto-Refresh Interval** - Configure periodic automatic usage queries with customizable intervals
- **Test Script API** - Validate JavaScript usage query scripts before execution
- **Enhanced Templates** - Custom blank templates with access token and user ID parameter support
Thanks to @Sirhexs
### Custom Configuration Directory (Cloud Sync)
- **Customizable Storage Location** - Customize CC Switch's configuration storage directory
- **Cloud Sync Support** - Point to cloud sync folders (Dropbox, OneDrive, iCloud Drive, etc.) to enable automatic config synchronization across devices
- **Independent Management** - Managed via Tauri Store for better isolation and reliability
Thanks to @ZyphrZero
### Configuration Directory Switching (WSL Support)
- **Auto-Sync on Directory Change** - When switching Claude/Codex config directories (e.g., WSL environment), automatically sync current provider to the new directory without manual operation
- **Post-Change Sync Utility** - Unified `postChangeSync.ts` utility for graceful error handling without blocking main flow
- **Import Config Auto-Sync** - Automatically sync after config import to ensure immediate effectiveness
- **Smart Conflict Resolution** - Distinguishes "fully successful" and "partially successful" states for precise user feedback
### Configuration Editor Improvements
- **JSON Format Button** - One-click JSON formatting in configuration editors
- **Real-Time TOML Validation** - Live syntax validation for Codex configuration with error highlighting
### Load Live Config When Editing
- **Protect Manual Modifications** - When editing the currently active provider, prioritize displaying the actual effective configuration from live files
- **Dual-Source Strategy** - Automatically loads from live config for active provider, SSOT for inactive ones
### Claude Configuration Data Structure Enhancements
- **Granular Model Configuration** - Migrated from dual-key to quad-key system for better model tier differentiation
- New fields: `ANTHROPIC_DEFAULT_HAIKU_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, `ANTHROPIC_DEFAULT_OPUS_MODEL`, `ANTHROPIC_MODEL`
- Replaces legacy `ANTHROPIC_SMALL_FAST_MODEL` with automatic migration
- Backend normalizes old configs on first read/write with smart fallback chain
- UI expanded from 2 to 4 model input fields with intelligent defaults
- **ANTHROPIC_API_KEY Support** - Providers can now use `ANTHROPIC_API_KEY` field in addition to `ANTHROPIC_AUTH_TOKEN`
- **Template Variable System** - Support for dynamic configuration replacement (e.g., KAT-Coder's `ENDPOINT_ID` parameter)
- **Endpoint Candidates** - Predefined endpoint list for speed testing and endpoint management
- **Visual Theme Configuration** - Custom icons and colors for provider cards
### Updated Provider Models
- **Kimi k2** - Updated to latest `kimi-k2-thinking` model
### New Provider Presets
Added 5 new provider presets:
- **DMXAPI** - Multi-model aggregation service
- **Azure Codex** - Microsoft Azure OpenAI endpoint
- **AnyRouter** - None-profit routing service
- **AiHubMix** - Multi-model aggregation service
- **MiniMax** - Open source AI model provider
### Partner Promotion Mechanism
- Support for ecosystem partner promotion (Zhipu GLM Z.ai)
- Sponsored banner integration in README
---
## Improvements
### Configuration & Sync
- **Unified Error Handling** - AppError with internationalized error messages throughout backend
- **Fixed apiKeyUrl Priority** - Correct priority order for API key URL resolution
- **Fixed MCP Sync Issues** - Resolved sync-to-other-side functionality failures
- **Import Config Sync** - Fixed sync issues after configuration import
- **Config Error Handling** - Force exit on config error to prevent silent fallback and data loss
### UI/UX Enhancements
- **Unique Provider Icons** - Each provider card now has unique icons and color identification
- **Unified Border System** - Consistent border design across all components
- **Drag Interaction** - Push effect animation and improved drag handle icons
- **Enhanced Visual Feedback** - Better current provider visual indication
- **Dialog Standardization** - Unified dialog sizes and layout consistency
- **Form Improvements** - Optimized model placeholders, simplified provider hints, category-specific hints
- **Usage Display Inline** - Usage info moved next to enable button for better space utilization
### Complete Internationalization
- **Error Messages i18n** - All backend error messages support Chinese/English
- **Tray Menu i18n** - System tray menu fully internationalized
- **UI Components i18n** - 100% coverage across all user-facing components
---
## Bug Fixes
### Configuration Management
- Fixed `apiKeyUrl` priority issue
- Fixed MCP sync-to-other-side functionality failure
- Fixed sync issues after config import
- Fixed Codex API Key auto-sync
- Fixed endpoint speed test functionality
- Fixed provider duplicate insertion position (now inserts next to original)
- Fixed custom endpoint preservation in edit mode
- Prevent silent fallback and data loss on config error
### Usage Query
- Fixed auto-query interval timing issue
- Ensured refresh button shows loading animation on click
### UI Issues
- Fixed name collision error (`get_init_error` command)
- Fixed language setting rollback after successful save
- Fixed language switch state reset (dependency cycle)
- Fixed edit mode button alignment
### Startup Issues
- Force exit on config error (no silent fallback)
- Eliminated code duplication causing initialization errors
---
## Architecture Refactoring
### Backend (Rust) - 5 Phase Refactoring
1. **Phase 1**: Unified error handling (`AppError` + i18n error messages)
2. **Phase 2**: Command layer split by domain (`commands/{provider,mcp,config,settings,plugin,misc}.rs`)
3. **Phase 3**: Integration tests and transaction mechanism (config snapshot + failure rollback)
4. **Phase 4**: Extracted Service layer (`services/{provider,mcp,config,speedtest}.rs`)
5. **Phase 5**: Concurrency optimization (`RwLock` instead of `Mutex`, scoped guard to avoid deadlock)
### Frontend (React + TypeScript) - 4 Stage Refactoring
1. **Stage 1**: Test infrastructure (vitest + MSW + @testing-library/react)
2. **Stage 2**: Extracted custom hooks (`useProviderActions`, `useMcpActions`, `useSettings`, `useImportExport`, etc.)
3. **Stage 3**: Component splitting and business logic extraction
4. **Stage 4**: Code cleanup and formatting unification
### Testing System
- **Hooks Unit Tests** - 100% coverage for all custom hooks
- **Integration Tests** - Coverage for key processes (App, SettingsDialog, MCP Panel)
- **MSW Mocking** - Backend API mocking to ensure test independence
- **Test Infrastructure** - vitest + MSW + @testing-library/react
### Code Quality
- **Unified Parameter Format** - All Tauri commands migrated to camelCase (Tauri 2 specification)
- **Semantic Clarity** - `AppType` renamed to `AppId` for better semantics
- **Centralized Parsing** - Unified `app` parameter parsing with `FromStr` trait
- **DRY Violations Cleanup** - Eliminated code duplication throughout codebase
- **Dead Code Removal** - Removed unused `missing_param` helper, deprecated `tauri-api.ts`, redundant `KimiModelSelector`
---
## Internal Optimizations (User Transparent)
### Removed Legacy Migration Logic
v3.6.0 removed v1 config auto-migration and copy file scanning logic:
- **Impact**: Improved startup performance, cleaner codebase
- **Compatibility**: v2 format configs fully compatible, no action required
- **Note**: Users upgrading from v3.1.0 or earlier should first upgrade to v3.2.x or v3.5.x for one-time migration, then upgrade to v3.6.0
### Command Parameter Standardization
Backend unified to use `app` parameter (values: `claude` or `codex`):
- **Impact**: More standardized code, friendlier error prompts
- **Compatibility**: Frontend fully adapted, users don't need to care about this change
---
## Dependencies
- Updated to **Tauri 2.8.x**
- Updated to **TailwindCSS 4.x**
- Updated to **TanStack Query v5.90.x**
- Maintained **React 18.2.x** and **TypeScript 5.3.x**
---
## Installation
### macOS
**Via Homebrew (Recommended):**
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
**Manual Download:**
- Download `CC-Switch-v3.6.0-macOS.zip` from [Assets](#assets) below
> **Note**: Due to lack of Apple Developer account, you may see "unidentified developer" warning. Go to System Settings → Privacy & Security → Click "Open Anyway"
### Windows
- **Installer**: `CC-Switch-v3.6.0-Windows.msi`
- **Portable**: `CC-Switch-v3.6.0-Windows-Portable.zip`
### Linux
- **AppImage**: `CC-Switch-v3.6.0-Linux.AppImage`
- **Debian**: `CC-Switch-v3.6.0-Linux.deb`
---
## Documentation
- [中文文档 (Chinese)](https://github.com/farion1231/cc-switch/blob/main/README_ZH.md)
- [English Documentation](https://github.com/farion1231/cc-switch/blob/main/README.md)
- [完整更新日志 (Full Changelog)](https://github.com/farion1231/cc-switch/blob/main/CHANGELOG.md)
---
## Acknowledgments
Special thanks to **Zhipu AI** for sponsoring this project with their GLM CODING PLAN!
---
**Full Changelog**: https://github.com/farion1231/cc-switch/compare/v3.5.1...v3.6.0
+249
View File
@@ -0,0 +1,249 @@
# CC Switch v3.6.0
> 全栈架构重构,增强配置同步与数据保护
**[English Version →](../release-note-v3.6.0.md)**
---
## 新增功能
### 编辑模式与供应商管理
- **供应商复制功能** - 一键快速复制现有供应商配置,轻松创建变体配置
- **手动排序功能** - 通过拖拽对供应商进行重新排序,带有视觉推送效果动画
- **编辑模式切换** - 显示/隐藏拖拽手柄,优化编辑体验
### 自定义端点管理
- **多端点配置** - 支持聚合类供应商的多 API 端点配置
- **端点输入可见性** - 为所有非官方供应商自动显示端点字段
### 自定义配置目录(云同步)
- **自定义存储位置** - 自定义 CC Switch 的配置存储目录
- **云同步支持** - 指定到云同步文件夹(Dropbox、OneDrive、iCloud Drive、坚果云等)即可实现跨设备配置自动同步
- **独立管理** - 通过 Tauri Store 管理,更好的隔离性和可靠性
### 使用量查询增强
- **自动刷新间隔** - 配置定时自动使用量查询,支持自定义间隔时间
- **测试脚本 API** - 在执行前验证 JavaScript 使用量查询脚本
- **增强模板系统** - 自定义空白模板,支持 access token 和 user ID 参数
### 配置目录切换(WSL 支持)
- **目录变更自动同步** - 切换 Claude/Codex 配置目录(如 WSL 环境)时,自动同步当前供应商到新目录,无需手动操作
- **后置同步工具** - 统一的 `postChangeSync.ts` 工具,优雅处理错误而不阻塞主流程
- **导入配置自动同步** - 配置导入后自动同步,确保立即生效
- **智能冲突解决** - 区分"完全成功"和"部分成功"状态,提供精确的用户反馈
### 配置编辑器改进
- **JSON 格式化按钮** - 配置编辑器中一键 JSON 格式化
- **实时 TOML 验证** - Codex 配置的实时语法验证,带有错误高亮
### 编辑时加载 Live 配置
- **保护手动修改** - 编辑当前激活的供应商时,优先显示来自 live 文件的实际生效配置
- **双源策略** - 活动供应商自动从 live 配置加载,非活动供应商从 SSOT 加载
### Claude 配置数据结构增强
- **细粒度模型配置** - 从双键系统升级到四键系统,以匹配官方最新数据结构
- 新增字段:`ANTHROPIC_DEFAULT_HAIKU_MODEL``ANTHROPIC_DEFAULT_SONNET_MODEL``ANTHROPIC_DEFAULT_OPUS_MODEL``ANTHROPIC_MODEL`
- 替换旧版 `ANTHROPIC_SMALL_FAST_MODEL`,支持自动迁移
- 后端在首次读写时自动规范化旧配置,带有智能回退链
- UI 从 2 个模型输入字段扩展到 4 个,具有智能默认值
- **ANTHROPIC_API_KEY 支持** - 供应商现可使用 `ANTHROPIC_API_KEY` 字段(除 `ANTHROPIC_AUTH_TOKEN` 外)
- **模板变量系统** - 支持动态配置替换(如 KAT-Coder 的 `ENDPOINT_ID` 参数)
- **端点候选列表** - 预定义端点列表,用于速度测试和端点管理
- **视觉主题配置** - 供应商卡片自定义图标和颜色
### 供应商模型更新
- **Kimi k2** - 更新到最新的 `kimi-k2-thinking` 模型
### 新增供应商预设
新增 5 个供应商预设:
- **DMXAPI** - 多模型聚合服务
- **Azure Codex** - 微软 Azure OpenAI 端点
- **AnyRouter** - API 路由服务
- **AiHubMix** - AI 模型集合
- **MiniMax** - 国产 AI 模型提供商
### 合作伙伴推广机制
- 支持生态合作伙伴推广(智谱 GLM Z.ai)
- README 中集成赞助商横幅
---
## 改进优化
### 配置与同步
- **统一错误处理** - 后端全面使用 AppError 与国际化错误消息
- **修复 apiKeyUrl 优先级** - 修正 API key URL 解析的优先级顺序
- **修复 MCP 同步问题** - 解决同步到另一端功能失效的问题
- **导入配置同步** - 修复配置导入后的同步问题
- **配置错误处理** - 配置错误时强制退出,防止静默回退和数据丢失
### UI/UX 增强
- **独特的供应商图标** - 每个供应商卡片现在都有独特的图标和颜色识别
- **统一边框系统** - 所有组件采用一致的边框设计
- **拖拽交互** - 推送效果动画和改进的拖拽手柄图标
- **增强视觉反馈** - 更好的当前供应商视觉指示
- **对话框标准化** - 统一的对话框尺寸和布局一致性
- **表单改进** - 优化模型占位符,简化供应商提示,分类特定提示
- **使用量内联显示** - 使用量信息移至启用按钮旁边,更好地利用空间
### 完整国际化
- **错误消息国际化** - 所有后端错误消息支持中英文
- **托盘菜单国际化** - 系统托盘菜单完全国际化
- **UI 组件国际化** - 所有面向用户的组件 100% 覆盖
---
## Bug 修复
### 配置管理
- 修复 `apiKeyUrl` 优先级问题
- 修复 MCP 同步到另一端功能失效
- 修复配置导入后的同步问题
- 修复 Codex API Key 自动同步
- 修复端点速度测试功能
- 修复供应商复制插入位置(现在插入到原供应商旁边)
- 修复编辑模式下自定义端点保留问题
- 防止配置错误时的静默回退和数据丢失
### 使用量查询
- 修复自动查询间隔时间问题
- 确保刷新按钮点击时显示加载动画
### UI 问题
- 修复名称冲突错误(`get_init_error` 命令)
- 修复保存成功后语言设置回滚
- 修复语言切换状态重置(依赖循环)
- 修复编辑模式按钮对齐
### 启动问题
- 配置错误时强制退出(不再静默回退)
- 消除导致初始化错误的代码重复
---
## 架构重构
### 后端(Rust- 5 阶段重构
1. **阶段 1**:统一错误处理(`AppError` + 国际化错误消息)
2. **阶段 2**:命令层按领域拆分(`commands/{provider,mcp,config,settings,plugin,misc}.rs`
3. **阶段 3**:集成测试和事务机制(配置快照 + 失败回滚)
4. **阶段 4**:提取 Service 层(`services/{provider,mcp,config,speedtest}.rs`
5. **阶段 5**:并发优化(`RwLock` 替代 `Mutex`,作用域 guard 避免死锁)
### 前端(React + TypeScript- 4 阶段重构
1. **阶段 1**:测试基础设施(vitest + MSW + @testing-library/react
2. **阶段 2**:提取自定义 hooks`useProviderActions``useMcpActions``useSettings``useImportExport` 等)
3. **阶段 3**:组件拆分和业务逻辑提取
4. **阶段 4**:代码清理和格式化统一
### 测试体系
- **Hooks 单元测试** - 所有自定义 hooks 100% 覆盖
- **集成测试** - 关键流程覆盖(App、SettingsDialog、MCP 面板)
- **MSW 模拟** - 后端 API 模拟确保测试独立性
- **测试基础设施** - vitest + MSW + @testing-library/react
### 代码质量
- **统一参数格式** - 所有 Tauri 命令迁移到 camelCaseTauri 2 规范)
- **语义清晰** - `AppType` 重命名为 `AppId` 以获得更好的语义
- **集中解析** - 使用 `FromStr` trait 统一 `app` 参数解析
- **DRY 违规清理** - 消除整个代码库中的代码重复
- **死代码移除** - 移除未使用的 `missing_param` 辅助函数、废弃的 `tauri-api.ts`、冗余的 `KimiModelSelector`
---
## 内部优化(用户无感知)
### 移除遗留迁移逻辑
v3.6.0 移除了 v1 配置自动迁移和副本文件扫描逻辑:
- **影响**:提升启动性能,代码更简洁
- **兼容性**:v2 格式配置完全兼容,无需任何操作
- **注意**:从 v3.1.0 或更早版本升级的用户,请先升级到 v3.2.x 或 v3.5.x 进行一次性迁移,然后再升级到 v3.6.0
### 命令参数标准化
后端统一使用 `app` 参数(取值:`claude``codex`):
- **影响**:代码更规范,错误提示更友好
- **兼容性**:前端已完全适配,用户无需关心此变更
---
## 依赖更新
- 更新到 **Tauri 2.8.x**
- 更新到 **TailwindCSS 4.x**
- 更新到 **TanStack Query v5.90.x**
- 保持 **React 18.2.x****TypeScript 5.3.x**
---
## 安装方式
### macOS
**通过 Homebrew 安装(推荐):**
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
**手动下载:**
- 从下方 [Assets](#assets) 下载 `CC-Switch-v3.6.0-macOS.zip`
> **注意**:由于作者没有苹果开发者账号,首次打开可能出现"未知开发者"警告。请前往"系统设置" → "隐私与安全性" → 点击"仍要打开"
### Windows
- **安装包**`CC-Switch-v3.6.0-Windows.msi`
- **便携版**`CC-Switch-v3.6.0-Windows-Portable.zip`
### Linux
- **AppImage**`CC-Switch-v3.6.0-Linux.AppImage`
- **Debian**`CC-Switch-v3.6.0-Linux.deb`
---
## 文档
- [中文文档](https://github.com/farion1231/cc-switch/blob/main/README_ZH.md)
- [English Documentation](https://github.com/farion1231/cc-switch/blob/main/README.md)
- [完整更新日志](https://github.com/farion1231/cc-switch/blob/main/CHANGELOG.md)
---
## 致谢
特别感谢**智谱 AI** 通过 GLM CODING PLAN 赞助本项目!
---
**完整变更记录**: https://github.com/farion1231/cc-switch/compare/v3.5.1...v3.6.0
+391
View File
@@ -0,0 +1,391 @@
# CC Switch v3.6.1
> Stability improvements and user experience optimization (based on v3.6.0)
**[中文更新说明 Chinese Documentation →](https://github.com/farion1231/cc-switch/blob/main/docs/release-note-v3.6.1-zh.md)**
---
## 📦 What's New in v3.6.1 (2025-11-10)
This release focuses on **user experience optimization** and **configuration parsing robustness**, fixing several critical bugs and enhancing the usage query system.
### ✨ New Features
#### Usage Query System Enhancements
- **Credential Decoupling** - Usage queries can now use independent API Key and Base URL, no longer dependent on provider configuration
- Support for different query endpoints and authentication methods
- Automatically displays credential input fields based on template type
- General template: API Key + Base URL
- NewAPI template: Base URL + Access Token + User ID
- Custom template: Fully customizable
- **UI Component Upgrade** - Replaced native checkbox with shadcn/ui Switch component for modern experience
- **Form Unification** - Unified use of shadcn/ui Input components, consistent styling with the application
- **Password Visibility Toggle** - Added show/hide password functionality (API Key, Access Token)
#### Form Validation Infrastructure
- **Common Schema Library** - New JSON/TOML generic validators to reduce code duplication
- `jsonConfigSchema`: Generic JSON object validator
- `tomlConfigSchema`: Generic TOML format validator
- `mcpJsonConfigSchema`: MCP-specific JSON validator
- **MCP Conditional Field Validation** - Strict type checking
- stdio type requires `command` field
- http type requires `url` field
#### Partner Integration
- **PackyCode** - New official partner
- Added to Claude and Codex provider presets
- 10% discount promotion support
- New logo and partner identification
---
### 🔧 Improvements
#### User Experience
- **Drag Sort Sync** - Tray menu order now syncs with drag-and-drop sorting in real-time
- **Enhanced Error Notifications** - Provider switch failures now display copyable error messages
- **Removed Misleading Placeholders** - Deleted example text from model input fields to avoid user confusion
- **Auto-fill Base URL** - All non-official provider categories automatically populate the Base URL input field
#### Configuration Parsing
- **CJK Quote Normalization** - Automatically handles IME-input fullwidth quotes to prevent TOML parsing errors
- Supports automatic conversion of Chinese quotes (" " ' ') to ASCII quotes
- Applied in TOML input handlers
- Disabled browser auto-correction in Textarea component
- **Preserve Custom Fields** - Editing Codex MCP TOML configuration now preserves unknown fields
- Supports extension fields like timeout_ms, retry_count
- Forward compatibility with future MCP protocol extensions
---
### 🐛 Bug Fixes
#### Critical Fixes
- **Fixed usage script panel white screen crash** - FormLabel component missing FormField context caused entire app to crash
- Replaced with standalone Label component
- Root cause: FormLabel internally calls useFormField() hook which requires FormFieldContext
- **Fixed CJK input quote parsing failure** - IME-input fullwidth quotes caused TOML parsing errors
- Added textNormalization utility function
- Automatically normalizes quotes before parsing
- **Fixed drag sort tray desync** (#179) - Tray menu order not updated after drag-and-drop sorting
- Automatically calls updateTrayMenu after sorting completes
- Ensures UI and tray menu stay consistent
- **Fixed MCP custom field loss** - Custom fields silently dropped when editing Codex MCP configuration
- Uses spread operator to retain all fields
- Preserves unknown fields in normalizeServerConfig
#### Stability Improvements
- **Error Isolation** - Tray menu update failures no longer affect main operations
- Decoupled tray update errors from main operations
- Provides warning when main operation succeeds but tray update fails
- **Safe Pattern Matching** - Replaced `unwrap()` with safe pattern matching
- Avoids panic-induced app crashes
- Tray menu event handling uses match patterns
- **Import Config Classification** - Importing from default config now automatically sets category to `custom`
- Avoids imported configs being mistaken for official presets
- Provides clearer configuration source identification
---
### 📊 Technical Statistics
```
Commits: 17 commits
Code Changes: 31 files
- Additions: 1,163 lines
- Deletions: 811 lines
- Net Growth: +352 lines
Contributors: Jason (16), ZyphrZero (1)
```
**By Module**:
- UI/User Interface: 3 commits
- Usage Query System: 3 commits
- Configuration Parsing: 2 commits
- Form Validation: 1 commit
- Other Improvements: 8 commits
---
### 📥 Installation
#### macOS
**Via Homebrew (Recommended):**
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
**Manual Download:**
- Download `CC-Switch-v3.6.1-macOS.zip` from [Assets](#assets) below
> **Note**: Due to lack of Apple Developer account, you may see "unidentified developer" warning. Go to System Settings → Privacy & Security → Click "Open Anyway"
#### Windows
- **Installer**: `CC-Switch-v3.6.1-Windows.msi`
- **Portable**: `CC-Switch-v3.6.1-Windows-Portable.zip`
#### Linux
- **AppImage**: `CC-Switch-v3.6.1-Linux.AppImage`
- **Debian**: `CC-Switch-v3.6.1-Linux.deb`
---
### 📚 Documentation
- [中文文档 (Chinese)](https://github.com/farion1231/cc-switch/blob/main/README_ZH.md)
- [English Documentation](https://github.com/farion1231/cc-switch/blob/main/README.md)
- [完整更新日志 (Full Changelog)](https://github.com/farion1231/cc-switch/blob/main/CHANGELOG.md)
---
### 🙏 Acknowledgments
Special thanks to:
- **Zhipu AI** - For sponsoring this project with GLM CODING PLAN
- **PackyCode** - New official partner
- **ZyphrZero** - For contributing tray menu sync fix (#179)
---
**Full Changelog**: https://github.com/farion1231/cc-switch/compare/v3.6.0...v3.6.1
---
---
## 📜 v3.6.0 Complete Feature Review
> Content below is from v3.6.0 (2025-11-07), helping you understand the complete feature set
<details>
<summary><b>Click to expand v3.6.0 detailed content →</b></summary>
## What's New
### Edit Mode & Provider Management
- **Provider Duplication** - Quickly duplicate existing provider configurations to create variants with one click
- **Manual Sorting** - Drag and drop to reorder providers, with visual push effect animations. Thanks to @ZyphrZero
- **Edit Mode Toggle** - Show/hide drag handles to optimize editing experience
### Custom Endpoint Management
- **Multi-Endpoint Configuration** - Support for aggregator providers with multiple API endpoints
- **Endpoint Input Visibility** - Shows endpoint field for all non-official providers automatically
### Usage Query Enhancements
- **Auto-Refresh Interval** - Configure periodic automatic usage queries with customizable intervals
- **Test Script API** - Validate JavaScript usage query scripts before execution
- **Enhanced Templates** - Custom blank templates with access token and user ID parameter support
Thanks to @Sirhexs
### Custom Configuration Directory (Cloud Sync)
- **Customizable Storage Location** - Customize CC Switch's configuration storage directory
- **Cloud Sync Support** - Point to cloud sync folders (Dropbox, OneDrive, iCloud Drive, etc.) to enable automatic config synchronization across devices
- **Independent Management** - Managed via Tauri Store for better isolation and reliability
Thanks to @ZyphrZero
### Configuration Directory Switching (WSL Support)
- **Auto-Sync on Directory Change** - When switching Claude/Codex config directories (e.g., WSL environment), automatically sync current provider to the new directory without manual operation
- **Post-Change Sync Utility** - Unified `postChangeSync.ts` utility for graceful error handling without blocking main flow
- **Import Config Auto-Sync** - Automatically sync after config import to ensure immediate effectiveness
- **Smart Conflict Resolution** - Distinguishes "fully successful" and "partially successful" states for precise user feedback
### Configuration Editor Improvements
- **JSON Format Button** - One-click JSON formatting in configuration editors
- **Real-Time TOML Validation** - Live syntax validation for Codex configuration with error highlighting
### Load Live Config When Editing
- **Protect Manual Modifications** - When editing the currently active provider, prioritize displaying the actual effective configuration from live files
- **Dual-Source Strategy** - Automatically loads from live config for active provider, SSOT for inactive ones
### Claude Configuration Data Structure Enhancements
- **Granular Model Configuration** - Migrated from dual-key to quad-key system for better model tier differentiation
- New fields: `ANTHROPIC_DEFAULT_HAIKU_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, `ANTHROPIC_DEFAULT_OPUS_MODEL`, `ANTHROPIC_MODEL`
- Replaces legacy `ANTHROPIC_SMALL_FAST_MODEL` with automatic migration
- Backend normalizes old configs on first read/write with smart fallback chain
- UI expanded from 2 to 4 model input fields with intelligent defaults
- **ANTHROPIC_API_KEY Support** - Providers can now use `ANTHROPIC_API_KEY` field in addition to `ANTHROPIC_AUTH_TOKEN`
- **Template Variable System** - Support for dynamic configuration replacement (e.g., KAT-Coder's `ENDPOINT_ID` parameter)
- **Endpoint Candidates** - Predefined endpoint list for speed testing and endpoint management
- **Visual Theme Configuration** - Custom icons and colors for provider cards
### Updated Provider Models
- **Kimi k2** - Updated to latest `kimi-k2-thinking` model
### New Provider Presets
Added 5 new provider presets:
- **DMXAPI** - Multi-model aggregation service
- **Azure Codex** - Microsoft Azure OpenAI endpoint
- **AnyRouter** - None-profit routing service
- **AiHubMix** - Multi-model aggregation service
- **MiniMax** - Open source AI model provider
### Partner Promotion Mechanism
- Support for ecosystem partner promotion (Zhipu GLM Z.ai)
- Sponsored banner integration in README
---
## Improvements
### Configuration & Sync
- **Unified Error Handling** - AppError with internationalized error messages throughout backend
- **Fixed apiKeyUrl Priority** - Correct priority order for API key URL resolution
- **Fixed MCP Sync Issues** - Resolved sync-to-other-side functionality failures
- **Import Config Sync** - Fixed sync issues after configuration import
- **Config Error Handling** - Force exit on config error to prevent silent fallback and data loss
### UI/UX Enhancements
- **Unique Provider Icons** - Each provider card now has unique icons and color identification
- **Unified Border System** - Consistent border design across all components
- **Drag Interaction** - Push effect animation and improved drag handle icons
- **Enhanced Visual Feedback** - Better current provider visual indication
- **Dialog Standardization** - Unified dialog sizes and layout consistency
- **Form Improvements** - Optimized model placeholders, simplified provider hints, category-specific hints
- **Usage Display Inline** - Usage info moved next to enable button for better space utilization
### Complete Internationalization
- **Error Messages i18n** - All backend error messages support Chinese/English
- **Tray Menu i18n** - System tray menu fully internationalized
- **UI Components i18n** - 100% coverage across all user-facing components
---
## Bug Fixes
### Configuration Management
- Fixed `apiKeyUrl` priority issue
- Fixed MCP sync-to-other-side functionality failure
- Fixed sync issues after config import
- Fixed Codex API Key auto-sync
- Fixed endpoint speed test functionality
- Fixed provider duplicate insertion position (now inserts next to original)
- Fixed custom endpoint preservation in edit mode
- Prevent silent fallback and data loss on config error
### Usage Query
- Fixed auto-query interval timing issue
- Ensured refresh button shows loading animation on click
### UI Issues
- Fixed name collision error (`get_init_error` command)
- Fixed language setting rollback after successful save
- Fixed language switch state reset (dependency cycle)
- Fixed edit mode button alignment
### Startup Issues
- Force exit on config error (no silent fallback)
- Eliminated code duplication causing initialization errors
---
## Architecture Refactoring
### Backend (Rust) - 5 Phase Refactoring
1. **Phase 1**: Unified error handling (`AppError` + i18n error messages)
2. **Phase 2**: Command layer split by domain (`commands/{provider,mcp,config,settings,plugin,misc}.rs`)
3. **Phase 3**: Integration tests and transaction mechanism (config snapshot + failure rollback)
4. **Phase 4**: Extracted Service layer (`services/{provider,mcp,config,speedtest}.rs`)
5. **Phase 5**: Concurrency optimization (`RwLock` instead of `Mutex`, scoped guard to avoid deadlock)
### Frontend (React + TypeScript) - 4 Stage Refactoring
1. **Stage 1**: Test infrastructure (vitest + MSW + @testing-library/react)
2. **Stage 2**: Extracted custom hooks (`useProviderActions`, `useMcpActions`, `useSettings`, `useImportExport`, etc.)
3. **Stage 3**: Component splitting and business logic extraction
4. **Stage 4**: Code cleanup and formatting unification
### Testing System
- **Hooks Unit Tests** - 100% coverage for all custom hooks
- **Integration Tests** - Coverage for key processes (App, SettingsDialog, MCP Panel)
- **MSW Mocking** - Backend API mocking to ensure test independence
- **Test Infrastructure** - vitest + MSW + @testing-library/react
### Code Quality
- **Unified Parameter Format** - All Tauri commands migrated to camelCase (Tauri 2 specification)
- **Semantic Clarity** - `AppType` renamed to `AppId` for better semantics
- **Centralized Parsing** - Unified `app` parameter parsing with `FromStr` trait
- **DRY Violations Cleanup** - Eliminated code duplication throughout codebase
- **Dead Code Removal** - Removed unused `missing_param` helper, deprecated `tauri-api.ts`, redundant `KimiModelSelector`
---
## Internal Optimizations (User Transparent)
### Removed Legacy Migration Logic
v3.6.0 removed v1 config auto-migration and copy file scanning logic:
- **Impact**: Improved startup performance, cleaner codebase
- **Compatibility**: v2 format configs fully compatible, no action required
- **Note**: Users upgrading from v3.1.0 or earlier should first upgrade to v3.2.x or v3.5.x for one-time migration, then upgrade to v3.6.0
### Command Parameter Standardization
Backend unified to use `app` parameter (values: `claude` or `codex`):
- **Impact**: More standardized code, friendlier error prompts
- **Compatibility**: Frontend fully adapted, users don't need to care about this change
---
## Dependencies
- Updated to **Tauri 2.8.x**
- Updated to **TailwindCSS 4.x**
- Updated to **TanStack Query v5.90.x**
- Maintained **React 18.2.x** and **TypeScript 5.3.x**
</details>
---
## 🌟 About CC Switch
CC Switch is a cross-platform desktop application for managing and switching between different provider configurations for Claude Code and Codex. Built with Tauri 2.0 + React 18 + TypeScript, supporting Windows, macOS, and Linux.
**Core Features**:
- 🔄 One-click switching between multiple AI providers
- 📦 Support for both Claude Code and Codex applications
- 🎨 Modern UI with complete Chinese/English internationalization
- 🔐 Local storage, secure and reliable data
- ☁️ Support for cloud sync configurations
- 🧩 Unified MCP server management
---
**Project Repository**: https://github.com/farion1231/cc-switch
+389
View File
@@ -0,0 +1,389 @@
# CC Switch v3.6.1
> 稳定性提升与用户体验优化(基于 v3.6.0)
**[English Version →](../release-note-v3.6.1.md)**
---
## 📦 v3.6.1 新增内容 (2025-11-10)
本次更新主要聚焦于**用户体验优化**和**配置解析健壮性**,修复了多个关键 Bug,并增强了用量查询系统。
### ✨ 新增功能
#### 用量查询系统增强
- **凭证解耦** - 用量查询可使用独立的 API Key 和 Base URL,不再依赖供应商配置
- 支持不同的查询端点和认证方式
- 根据模板类型自动显示对应的凭证输入框
- General 模板:API Key + Base URL
- NewAPI 模板:Base URL + Access Token + User ID
- Custom 模板:完全自定义
- **UI 组件升级** - 使用 shadcn/ui Switch 替代原生 checkbox,体验更现代
- **表单统一化** - 统一使用 shadcn/ui 输入组件,样式与应用保持一致
- **密码显示切换** - 添加查看/隐藏密码功能(API Key、Access Token
#### 表单验证基础设施
- **通用 Schema 库** - 新增 JSON/TOML 通用验证器,减少重复代码
- `jsonConfigSchema`:通用 JSON 对象验证器
- `tomlConfigSchema`:通用 TOML 格式验证器
- `mcpJsonConfigSchema`MCP 专用 JSON 验证器
- **MCP 条件字段验证** - 严格的类型检查
- stdio 类型强制要求 `command` 字段
- http 类型强制要求 `url` 字段
#### 合作伙伴集成
- **PackyCode** - 新增官方合作伙伴
- 添加到 Claude 和 Codex 供应商预设
- 支持 10% 折扣优惠(促销信息集成)
- 新增 Logo 和合作伙伴标识
---
### 🔧 改进优化
#### 用户体验
- **拖拽排序同步** - 托盘菜单顺序实时同步拖拽排序结果
- **错误通知增强** - 切换供应商失败时显示可复制的错误信息
- **移除误导性占位符** - 删除模型输入框的示例文本,避免用户混淆
- **Base URL 自动填充** - 所有非官方供应商类别自动填充 Base URL 输入框
#### 配置解析
- **中文引号规范化** - 自动处理 IME 输入的全角引号,防止 TOML 解析错误
- 支持中文引号(" " ' ')自动转换为 ASCII 引号
- 在 TOML 输入处理器中应用
- Textarea 组件禁用浏览器自动纠正
- **自定义字段保留** - 编辑 Codex MCP TOML 配置时保留未知字段
- 支持 timeout_ms、retry_count 等扩展字段
- 向前兼容未来的 MCP 协议扩展
---
### 🐛 Bug 修复
#### 关键修复
- **修复用量脚本面板白屏崩溃** - FormLabel 组件缺少 FormField context 导致整个应用崩溃
- 替换为独立的 Label 组件
- 根本原因:FormLabel 内部调用 useFormField() hook 需要 FormFieldContext
- **修复中文输入法引号解析失败** - IME 输入的全角引号导致 TOML 解析错误
- 新增 textNormalization 工具函数
- 在解析前自动规范化引号
- **修复拖拽排序托盘不同步** (#179) - 拖拽排序后托盘菜单顺序未更新
- 在排序完成后自动调用 updateTrayMenu
- 确保 UI 和托盘菜单保持一致
- **修复 MCP 自定义字段丢失** - 编辑 Codex MCP 配置时自定义字段被静默丢弃
- 使用 spread 操作符保留所有字段
- normalizeServerConfig 中保留未知字段
#### 稳定性改进
- **错误隔离** - 托盘菜单更新失败不再影响主操作流程
- 将托盘更新错误与主操作解耦
- 主操作成功但托盘更新失败时给出警告
- **安全模式匹配** - 替换 `unwrap()` 为安全的 pattern matching
- 避免 panic 导致应用崩溃
- 托盘菜单事件处理使用 match 模式
- **导入配置分类** - 从默认配置导入时自动设置 category 为 `custom`
- 避免导入的配置被误认为官方预设
- 提供更清晰的配置来源标识
---
### 📊 技术统计
```
提交数: 17 commits
代码变更: 31 个文件
- 新增: 1,163 行
- 删除: 811 行
- 净增长: +352 行
贡献者: Jason (16), ZyphrZero (1)
```
**按模块分类**
- UI/用户界面:3 commits
- 用量查询系统:3 commits
- 配置解析:2 commits
- 表单验证:1 commit
- 其他改进:8 commits
---
### 📥 安装方式
#### macOS
**通过 Homebrew 安装(推荐):**
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
**手动下载:**
- 从下方 [Assets](#assets) 下载 `CC-Switch-v3.6.1-macOS.zip`
> **注意**:由于作者没有苹果开发者账号,首次打开可能出现"未知开发者"警告。请前往"系统设置" → "隐私与安全性" → 点击"仍要打开"
#### Windows
- **安装包**`CC-Switch-v3.6.1-Windows.msi`
- **便携版**`CC-Switch-v3.6.1-Windows-Portable.zip`
#### Linux
- **AppImage**`CC-Switch-v3.6.1-Linux.AppImage`
- **Debian**`CC-Switch-v3.6.1-Linux.deb`
---
### 📚 文档
- [中文文档](https://github.com/farion1231/cc-switch/blob/main/README_ZH.md)
- [English Documentation](https://github.com/farion1231/cc-switch/blob/main/README.md)
- [完整更新日志](https://github.com/farion1231/cc-switch/blob/main/CHANGELOG.md)
---
### 🙏 致谢
特别感谢:
- **智谱 AI** - 通过 GLM CODING PLAN 赞助本项目
- **PackyCode** - 新加入的官方合作伙伴
- **ZyphrZero** - 贡献托盘菜单同步修复 (#179)
---
**完整变更记录**: https://github.com/farion1231/cc-switch/compare/v3.6.0...v3.6.1
---
---
## 📜 v3.6.0 完整功能回顾
> 以下内容来自 v3.6.0 (2025-11-07),帮助您了解完整的功能集
<details>
<summary><b>点击展开 v3.6.0 的详细内容 →</b></summary>
## 新增功能
### 编辑模式与供应商管理
- **供应商复制功能** - 一键快速复制现有供应商配置,轻松创建变体配置
- **手动排序功能** - 通过拖拽对供应商进行重新排序,带有视觉推送效果动画
- **编辑模式切换** - 显示/隐藏拖拽手柄,优化编辑体验
### 自定义端点管理
- **多端点配置** - 支持聚合类供应商的多 API 端点配置
- **端点输入可见性** - 为所有非官方供应商自动显示端点字段
### 自定义配置目录(云同步)
- **自定义存储位置** - 自定义 CC Switch 的配置存储目录
- **云同步支持** - 指定到云同步文件夹(Dropbox、OneDrive、iCloud Drive、坚果云等)即可实现跨设备配置自动同步
- **独立管理** - 通过 Tauri Store 管理,更好的隔离性和可靠性
### 使用量查询增强
- **自动刷新间隔** - 配置定时自动使用量查询,支持自定义间隔时间
- **测试脚本 API** - 在执行前验证 JavaScript 使用量查询脚本
- **增强模板系统** - 自定义空白模板,支持 access token 和 user ID 参数
### 配置目录切换(WSL 支持)
- **目录变更自动同步** - 切换 Claude/Codex 配置目录(如 WSL 环境)时,自动同步当前供应商到新目录,无需手动操作
- **后置同步工具** - 统一的 `postChangeSync.ts` 工具,优雅处理错误而不阻塞主流程
- **导入配置自动同步** - 配置导入后自动同步,确保立即生效
- **智能冲突解决** - 区分"完全成功"和"部分成功"状态,提供精确的用户反馈
### 配置编辑器改进
- **JSON 格式化按钮** - 配置编辑器中一键 JSON 格式化
- **实时 TOML 验证** - Codex 配置的实时语法验证,带有错误高亮
### 编辑时加载 Live 配置
- **保护手动修改** - 编辑当前激活的供应商时,优先显示来自 live 文件的实际生效配置
- **双源策略** - 活动供应商自动从 live 配置加载,非活动供应商从 SSOT 加载
### Claude 配置数据结构增强
- **细粒度模型配置** - 从双键系统升级到四键系统,以匹配官方最新数据结构
- 新增字段:`ANTHROPIC_DEFAULT_HAIKU_MODEL``ANTHROPIC_DEFAULT_SONNET_MODEL``ANTHROPIC_DEFAULT_OPUS_MODEL``ANTHROPIC_MODEL`
- 替换旧版 `ANTHROPIC_SMALL_FAST_MODEL`,支持自动迁移
- 后端在首次读写时自动规范化旧配置,带有智能回退链
- UI 从 2 个模型输入字段扩展到 4 个,具有智能默认值
- **ANTHROPIC_API_KEY 支持** - 供应商现可使用 `ANTHROPIC_API_KEY` 字段(除 `ANTHROPIC_AUTH_TOKEN` 外)
- **模板变量系统** - 支持动态配置替换(如 KAT-Coder 的 `ENDPOINT_ID` 参数)
- **端点候选列表** - 预定义端点列表,用于速度测试和端点管理
- **视觉主题配置** - 供应商卡片自定义图标和颜色
### 供应商模型更新
- **Kimi k2** - 更新到最新的 `kimi-k2-thinking` 模型
### 新增供应商预设
新增 5 个供应商预设:
- **DMXAPI** - 多模型聚合服务
- **Azure Codex** - 微软 Azure OpenAI 端点
- **AnyRouter** - API 路由服务
- **AiHubMix** - AI 模型集合
- **MiniMax** - 国产 AI 模型提供商
### 合作伙伴推广机制
- 支持生态合作伙伴推广(智谱 GLM Z.ai)
- README 中集成赞助商横幅
---
## 改进优化
### 配置与同步
- **统一错误处理** - 后端全面使用 AppError 与国际化错误消息
- **修复 apiKeyUrl 优先级** - 修正 API key URL 解析的优先级顺序
- **修复 MCP 同步问题** - 解决同步到另一端功能失效的问题
- **导入配置同步** - 修复配置导入后的同步问题
- **配置错误处理** - 配置错误时强制退出,防止静默回退和数据丢失
### UI/UX 增强
- **独特的供应商图标** - 每个供应商卡片现在都有独特的图标和颜色识别
- **统一边框系统** - 所有组件采用一致的边框设计
- **拖拽交互** - 推送效果动画和改进的拖拽手柄图标
- **增强视觉反馈** - 更好的当前供应商视觉指示
- **对话框标准化** - 统一的对话框尺寸和布局一致性
- **表单改进** - 优化模型占位符,简化供应商提示,分类特定提示
- **使用量内联显示** - 使用量信息移至启用按钮旁边,更好地利用空间
### 完整国际化
- **错误消息国际化** - 所有后端错误消息支持中英文
- **托盘菜单国际化** - 系统托盘菜单完全国际化
- **UI 组件国际化** - 所有面向用户的组件 100% 覆盖
---
## Bug 修复
### 配置管理
- 修复 `apiKeyUrl` 优先级问题
- 修复 MCP 同步到另一端功能失效
- 修复配置导入后的同步问题
- 修复 Codex API Key 自动同步
- 修复端点速度测试功能
- 修复供应商复制插入位置(现在插入到原供应商旁边)
- 修复编辑模式下自定义端点保留问题
- 防止配置错误时的静默回退和数据丢失
### 使用量查询
- 修复自动查询间隔时间问题
- 确保刷新按钮点击时显示加载动画
### UI 问题
- 修复名称冲突错误(`get_init_error` 命令)
- 修复保存成功后语言设置回滚
- 修复语言切换状态重置(依赖循环)
- 修复编辑模式按钮对齐
### 启动问题
- 配置错误时强制退出(不再静默回退)
- 消除导致初始化错误的代码重复
---
## 架构重构
### 后端(Rust- 5 阶段重构
1. **阶段 1**:统一错误处理(`AppError` + 国际化错误消息)
2. **阶段 2**:命令层按领域拆分(`commands/{provider,mcp,config,settings,plugin,misc}.rs`
3. **阶段 3**:集成测试和事务机制(配置快照 + 失败回滚)
4. **阶段 4**:提取 Service 层(`services/{provider,mcp,config,speedtest}.rs`
5. **阶段 5**:并发优化(`RwLock` 替代 `Mutex`,作用域 guard 避免死锁)
### 前端(React + TypeScript- 4 阶段重构
1. **阶段 1**:测试基础设施(vitest + MSW + @testing-library/react
2. **阶段 2**:提取自定义 hooks`useProviderActions``useMcpActions``useSettings``useImportExport` 等)
3. **阶段 3**:组件拆分和业务逻辑提取
4. **阶段 4**:代码清理和格式化统一
### 测试体系
- **Hooks 单元测试** - 所有自定义 hooks 100% 覆盖
- **集成测试** - 关键流程覆盖(App、SettingsDialog、MCP 面板)
- **MSW 模拟** - 后端 API 模拟确保测试独立性
- **测试基础设施** - vitest + MSW + @testing-library/react
### 代码质量
- **统一参数格式** - 所有 Tauri 命令迁移到 camelCaseTauri 2 规范)
- **语义清晰** - `AppType` 重命名为 `AppId` 以获得更好的语义
- **集中解析** - 使用 `FromStr` trait 统一 `app` 参数解析
- **DRY 违规清理** - 消除整个代码库中的代码重复
- **死代码移除** - 移除未使用的 `missing_param` 辅助函数、废弃的 `tauri-api.ts`、冗余的 `KimiModelSelector`
---
## 内部优化(用户无感知)
### 移除遗留迁移逻辑
v3.6.0 移除了 v1 配置自动迁移和副本文件扫描逻辑:
- **影响**:提升启动性能,代码更简洁
- **兼容性**:v2 格式配置完全兼容,无需任何操作
- **注意**:从 v3.1.0 或更早版本升级的用户,请先升级到 v3.2.x 或 v3.5.x 进行一次性迁移,然后再升级到 v3.6.0
### 命令参数标准化
后端统一使用 `app` 参数(取值:`claude``codex`):
- **影响**:代码更规范,错误提示更友好
- **兼容性**:前端已完全适配,用户无需关心此变更
---
## 依赖更新
- 更新到 **Tauri 2.8.x**
- 更新到 **TailwindCSS 4.x**
- 更新到 **TanStack Query v5.90.x**
- 保持 **React 18.2.x****TypeScript 5.3.x**
</details>
---
## 🌟 关于 CC Switch
CC Switch 是一个跨平台桌面应用,用于管理和切换 Claude Code 与 Codex 的不同供应商配置。基于 Tauri 2.0 + React 18 + TypeScript 构建,支持 Windows、macOS、Linux。
**核心特性**
- 🔄 一键切换多个 AI 供应商
- 📦 支持 Claude Code 和 Codex 双应用
- 🎨 现代化 UI,完整的中英文国际化
- 🔐 本地存储,数据安全可靠
- ☁️ 支持云同步配置
- 🧩 MCP 服务器统一管理
---
**项目地址**: https://github.com/farion1231/cc-switch
+439
View File
@@ -0,0 +1,439 @@
# CC Switch v3.7.0
> From Provider Switcher to All-in-One AI CLI Management Platform
**[中文更新说明 Chinese Documentation →](release-note-v3.7.0-zh.md)**
---
## Overview
CC Switch v3.7.0 introduces six major features with over 18,000 lines of new code.
**Release Date**: 2025-11-19
**Commits**: 85 from v3.6.0
**Code Changes**: 152 files, +18,104 / -3,732 lines
---
## New Features
### Gemini CLI Integration
Complete support for Google Gemini CLI, becoming the third supported application (Claude Code, Codex, Gemini).
**Core Capabilities**:
- **Dual-file configuration** - Support for both `.env` and `settings.json` formats
- **Auto-detection** - Automatically detect `GOOGLE_GEMINI_BASE_URL`, `GEMINI_MODEL`, etc.
- **Full MCP support** - Complete MCP server management for Gemini
- **Deep link integration** - Import via `ccswitch://` protocol
- **System tray** - Quick-switch from tray menu
**Provider Presets**:
- **Google Official** - OAuth authentication support
- **PackyCode** - Partner integration
- **Custom** - Full customization support
**Technical Implementation**:
- New backend modules: `gemini_config.rs` (20KB), `gemini_mcp.rs`
- Form synchronization with environment editor
- Dual-file atomic writes
---
### MCP v3.7.0 Unified Architecture
Complete refactoring of MCP management system for cross-application unification.
**Architecture Improvements**:
- **Unified panel** - Single interface for Claude/Codex/Gemini MCP servers
- **SSE transport** - New Server-Sent Events support
- **Smart parser** - Fault-tolerant JSON parsing
- **Format correction** - Auto-fix Codex `[mcp_servers]` format
- **Extended fields** - Preserve custom TOML fields
**User Experience**:
- Default app selection in forms
- JSON formatter for validation
- Improved visual hierarchy
- Better error messages
**Import/Export**:
- Unified import from all three apps
- Bidirectional synchronization
- State preservation
---
### Claude Skills Management System
**Approximately 2,000 lines of code** - A complete skill ecosystem platform.
**GitHub Integration**:
- Auto-scan skills from GitHub repositories
- Pre-configured repos:
- `ComposioHQ/awesome-claude-skills` - Curated collection
- `anthropics/skills` - Official Anthropic skills
- `cexll/myclaude` - Community contributions
- Add custom repositories
- Subdirectory scanning support (`skillsPath`)
**Lifecycle Management**:
- **Discover** - Auto-detect `SKILL.md` files
- **Install** - One-click to `~/.claude/skills/`
- **Uninstall** - Safe removal with tracking
- **Update** - Check for updates (infrastructure ready)
**Technical Architecture**:
- **Backend**: `SkillService` (526 lines) with GitHub API integration
- **Frontend**: SkillsPage, SkillCard, RepoManager
- **UI Components**: Badge, Card, Table (shadcn/ui)
- **State**: Persistent storage in `skills.json`
- **i18n**: 47+ translation keys
---
### Prompts Management System
**Approximately 1,300 lines of code** - Complete system prompt management.
**Multi-Preset Management**:
- Create unlimited prompt presets
- Quick switch between presets
- One active prompt at a time
- Delete protection for active prompts
**Cross-App Support**:
- **Claude**: `~/.claude/CLAUDE.md`
- **Codex**: `~/.codex/AGENTS.md`
- **Gemini**: `~/.gemini/GEMINI.md`
**Markdown Editor**:
- Full-featured CodeMirror 6 integration
- Syntax highlighting
- Dark theme (One Dark)
- Real-time preview
**Smart Synchronization**:
- **Auto-write** - Immediately write to live files
- **Backfill protection** - Save current content before switching
- **Auto-import** - Import from live files on first launch
- **Modification protection** - Preserve manual modifications
**Technical Implementation**:
- **Backend**: `PromptService` (213 lines)
- **Frontend**: PromptPanel (177), PromptFormModal (160), MarkdownEditor (159)
- **Hooks**: usePromptActions (152 lines)
- **i18n**: 41+ translation keys
---
### Deep Link Protocol (ccswitch://)
One-click provider configuration import via URL scheme.
**Features**:
- Protocol registration on all platforms
- Import from shared links
- Lifecycle integration
- Security validation
---
### Environment Variable Conflict Detection
Intelligent detection and management of configuration conflicts.
**Detection Scope**:
- **Claude & Codex** - Cross-app conflicts
- **Gemini** - Auto-discovery
- **MCP** - Server configuration conflicts
**Management Features**:
- Visual conflict indicators
- Resolution suggestions
- Override warnings
- Backup before changes
---
## Improvements
### Provider Management
**New Presets**:
- **DouBaoSeed** - ByteDance's DouBao
- **Kimi For Coding** - Moonshot AI
- **BaiLing** - BaiLing AI
- **Removed AnyRouter** - To avoid confusion
**Enhancements**:
- Model name configuration for Codex and Gemini
- Provider notes field for organization
- Enhanced preset metadata
### Configuration Management
- **Common config migration** - From localStorage to `config.json`
- **Unified persistence** - Shared across all apps
- **Auto-import** - First launch configuration import
- **Backfill priority** - Correct handling of live files
### UI/UX Improvements
**Design System**:
- **macOS native** - System-aligned color scheme
- **Window centering** - Default centered position
- **Visual polish** - Improved spacing and hierarchy
**Interactions**:
- **Password input** - Fixed Edge/IE reveal buttons
- **URL overflow** - Fixed card overflow
- **Error copying** - Copy-to-clipboard errors
- **Tray sync** - Real-time drag-and-drop sync
---
## Bug Fixes
### Critical Fixes
- **Usage script validation** - Boundary checks
- **Gemini validation** - Relaxed constraints
- **TOML parsing** - CJK quote handling
- **MCP fields** - Custom field preservation
- **White screen** - FormLabel crash fix
### Stability
- **Tray safety** - Pattern matching instead of unwrap
- **Error isolation** - Tray failures don't block operations
- **Import classification** - Correct category assignment
### UI Fixes
- **Model placeholders** - Removed misleading hints
- **Base URL** - Auto-fill for third-party providers
- **Drag sort** - Tray menu synchronization
---
## Technical Improvements
### Architecture
**MCP v3.7.0**:
- Removed legacy code (~1,000 lines)
- Unified initialization structure
- Backward compatibility maintained
- Comprehensive code formatting
**Platform Compatibility**:
- Windows winreg API fix (v0.52)
- Safe pattern matching (no `unwrap()`)
- Cross-platform tray handling
### Configuration
**Synchronization**:
- MCP sync across all apps
- Gemini form-editor sync
- Dual-file reading (.env + settings.json)
**Validation**:
- Input boundary checks
- TOML quote normalization (CJK)
- Custom field preservation
- Enhanced error messages
### Code Quality
**Type Safety**:
- Complete TypeScript coverage
- Rust type refinements
- API contract validation
**Testing**:
- Simplified assertions
- Better test coverage
- Integration test updates
**Dependencies**:
- Tauri 2.8.x
- Rust: `anyhow`, `zip`, `serde_yaml`, `tempfile`
- Frontend: CodeMirror 6 packages
- winreg 0.52 (Windows)
---
## Technical Statistics
```
Total Changes:
- Commits: 85
- Files: 152 changed
- Additions: +18,104 lines
- Deletions: -3,732 lines
New Modules:
- Skills Management: 2,034 lines (21 files)
- Prompts Management: 1,302 lines (20 files)
- Gemini Integration: ~1,000 lines
- MCP Refactor: ~3,000 lines refactored
Code Distribution:
- Backend (Rust): ~4,500 lines new
- Frontend (React): ~3,000 lines new
- Configuration: ~1,500 lines refactored
- Tests: ~500 lines
```
---
## Strategic Positioning
### From Tool to Platform
v3.7.0 represents a shift in CC Switch's positioning:
| Aspect | v3.6 | v3.7.0 |
| ----------------- | ------------------------ | ---------------------------- |
| **Identity** | Provider Switcher | AI CLI Management Platform |
| **Scope** | Configuration Management | Ecosystem Management |
| **Applications** | Claude + Codex | Claude + Codex + Gemini |
| **Capabilities** | Switch configs | Extend capabilities (Skills) |
| **Customization** | Manual editing | Visual management (Prompts) |
| **Integration** | Isolated apps | Unified management (MCP) |
### Six Pillars of AI CLI Management
1. **Configuration Management** - Provider switching and management
2. **Capability Extension** - Skills installation and lifecycle
3. **Behavior Customization** - System prompt presets
4. **Ecosystem Integration** - Deep links and sharing
5. **Multi-AI Support** - Claude/Codex/Gemini
6. **Intelligent Detection** - Conflict prevention
---
## Download & Installation
### System Requirements
- **Windows**: Windows 10+
- **macOS**: macOS 10.15 (Catalina)+
- **Linux**: Ubuntu 22.04+ / Debian 11+ / Fedora 34+
### Download Links
Visit [Releases](https://github.com/farion1231/cc-switch/releases/latest) to download:
- **Windows**: `CC-Switch-v3.7.0-Windows.msi` or `-Portable.zip`
- **macOS**: `CC-Switch-v3.7.0-macOS.tar.gz` or `.zip`
- **Linux**: `CC-Switch-v3.7.0-Linux.AppImage` or `.deb`
### Homebrew (macOS)
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
Update:
```bash
brew upgrade --cask cc-switch
```
---
## Migration Notes
### From v3.6.x
**Automatic migration** - No action required, configs are fully compatible
### From v3.1.x or Earlier
**Two-step migration required**:
1. First upgrade to v3.2.x (performs one-time migration)
2. Then upgrade to v3.7.0
### New Features
- **Skills**: No migration needed, start fresh
- **Prompts**: Auto-import from live files on first launch
- **Gemini**: Install Gemini CLI separately if needed
- **MCP v3.7.0**: Backward compatible with previous configs
---
## Acknowledgments
### Contributors
Thanks to all contributors who made this release possible:
- [@YoVinchen](https://github.com/YoVinchen) - Skills & Prompts & Gemini integration implementation
- [@farion1231](https://github.com/farion1231) - From developer to issue responder
- Community members for testing and feedback
### Sponsors
**Z.ai** - GLM CODING PLAN sponsor
[Get 10% OFF with this link](https://z.ai/subscribe?ic=8JVLJQFSKB)
**PackyCode** - API relay service partner
[Register with "cc-switch" code for 10% discount](https://www.packyapi.com/register?aff=cc-switch)
---
## Feedback & Support
- **Issues**: [GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **Discussions**: [GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **Documentation**: [README](../README.md)
- **Changelog**: [CHANGELOG.md](../CHANGELOG.md)
---
## What's Next
**v3.8.0 Preview** (Tentative):
- Local proxy functionality
Stay tuned for more updates!
---
**Happy Coding!**
+435
View File
@@ -0,0 +1,435 @@
# CC Switch v3.7.0
> 从供应商切换器到 AI CLI 一体化管理平台
**[English Version →](release-note-v3.7.0-en.md)**
---
## 概览
CC Switch v3.7.0 新增六大核心功能,新增超过 18,000 行代码。
**发布日期**2025-11-19
**提交数量**:从 v3.6.0 开始 85 个提交
**代码变更**152 个文件,+18,104 / -3,732 行
---
## 新增功能
### Gemini CLI 集成
完整支持 Google Gemini CLI,成为第三个支持的应用(Claude Code、Codex、Gemini)。
**核心能力**
- **双文件配置** - 同时支持 `.env``settings.json` 格式
- **自动检测** - 自动检测 `GOOGLE_GEMINI_BASE_URL``GEMINI_MODEL` 等环境变量
- **完整 MCP 支持** - 为 Gemini 提供完整的 MCP 服务器管理
- **深度链接集成** - 通过 `ccswitch://` 协议导入配置
- **系统托盘** - 从托盘菜单快速切换
**供应商预设**
- **Google Official** - 支持 OAuth 认证
- **PackyCode** - 合作伙伴集成
- **自定义** - 完全自定义支持
**技术实现**
- 新增后端模块:`gemini_config.rs`20KB)、`gemini_mcp.rs`
- 表单与环境编辑器同步
- 双文件原子写入
---
### MCP v3.7.0 统一架构
MCP 管理系统完整重构,实现跨应用统一管理。
**架构改进**
- **统一管理面板** - 单一界面管理 Claude/Codex/Gemini MCP 服务器
- **SSE 传输类型** - 新增 Server-Sent Events 支持
- **智能解析器** - 容错性 JSON 解析
- **格式修正** - 自动修复 Codex `[mcp_servers]` 格式
- **扩展字段** - 保留自定义 TOML 字段
**用户体验**
- 表单中的默认应用选择
- JSON 格式化器用于验证
- 改进的视觉层次
- 更好的错误消息
**导入/导出**
- 统一从三个应用导入
- 双向同步
- 状态保持
---
### Claude Skills 管理系统
**约 2,000 行代码** - 完整的技能生态平台。
**GitHub 集成**
- 从 GitHub 仓库自动扫描技能
- 预配置仓库:
- `ComposioHQ/awesome-claude-skills` - 精选集合
- `anthropics/skills` - Anthropic 官方技能
- `cexll/myclaude` - 社区贡献
- 添加自定义仓库
- 子目录扫描支持(`skillsPath`
**生命周期管理**
- **发现** - 自动检测 `SKILL.md` 文件
- **安装** - 一键安装到 `~/.claude/skills/`
- **卸载** - 安全移除并跟踪状态
- **更新** - 检查更新(基础设施已就绪)
**技术架构**
- **后端**`SkillService`526 行)集成 GitHub API
- **前端**SkillsPage、SkillCard、RepoManager
- **UI 组件**Badge、Card、Tableshadcn/ui
- **状态**:持久化存储在 `skills.json`
- **国际化**47+ 个翻译键
---
### Prompts 管理系统
**约 1,300 行代码** - 完整的系统提示词管理。
**多预设管理**
- 创建无限数量的提示词预设
- 快速在预设间切换
- 同时只能激活一个提示词
- 活动提示词删除保护
**跨应用支持**
- **Claude**`~/.claude/CLAUDE.md`
- **Codex**`~/.codex/AGENTS.md`
- **Gemini**`~/.gemini/GEMINI.md`
**Markdown 编辑器**
- 完整的 CodeMirror 6 集成
- 语法高亮
- 暗色主题(One Dark
- 实时预览
**智能同步**
- **自动写入** - 立即写入 live 文件
- **回填保护** - 切换前保存当前内容
- **自动导入** - 首次启动从 live 文件导入
- **修改保护** - 保留手动修改
**技术实现**
- **后端**`PromptService`213 行)
- **前端**PromptPanel177)、PromptFormModal160)、MarkdownEditor159
- **Hooks**usePromptActions152 行)
- **国际化**41+ 个翻译键
---
### 深度链接协议(ccswitch://
通过 URL 方案一键导入供应商配置。
**功能特性**
- 所有平台的协议注册
- 从共享链接导入
- 生命周期集成
- 安全验证
---
### 环境变量冲突检测
智能检测和管理配置冲突。
**检测范围**
- **Claude & Codex** - 跨应用冲突
- **Gemini** - 自动发现
- **MCP** - 服务器配置冲突
**管理功能**
- 可视化冲突指示器
- 解决建议
- 覆盖警告
- 更改前备份
---
## 改进优化
### 供应商管理
**新增预设**
- **DouBaoSeed** - 字节跳动的豆包
- **Kimi For Coding** - 月之暗面
- **BaiLing** - 百灵 AI
- **移除 AnyRouter** - 避免误导
**增强功能**
- Codex 和 Gemini 的模型名称配置
- 供应商备注字段用于组织
- 增强的预设元数据
### 配置管理
- **通用配置迁移** - 从 localStorage 迁移到 `config.json`
- **统一持久化** - 跨所有应用共享
- **自动导入** - 首次启动配置导入
- **回填优先级** - 正确处理 live 文件
### UI/UX 改进
**设计系统**
- **macOS 原生** - 与系统对齐的配色方案
- **窗口居中** - 默认居中位置
- **视觉优化** - 改进的间距和层次
**交互优化**
- **密码输入** - 修复 Edge/IE 显示按钮
- **URL 溢出** - 修复卡片溢出
- **错误复制** - 可复制到剪贴板的错误
- **托盘同步** - 实时拖放同步
---
## Bug 修复
### 关键修复
- **用量脚本验证** - 边界检查
- **Gemini 验证** - 放宽约束
- **TOML 解析** - CJK 引号处理
- **MCP 字段** - 自定义字段保留
- **白屏** - FormLabel 崩溃修复
### 稳定性
- **托盘安全** - 模式匹配替代 unwrap
- **错误隔离** - 托盘失败不阻塞操作
- **导入分类** - 正确的类别分配
### UI 修复
- **模型占位符** - 移除误导性提示
- **Base URL** - 第三方供应商自动填充
- **拖拽排序** - 托盘菜单同步
---
## 技术改进
### 架构
**MCP v3.7.0**
- 移除遗留代码(约 1,000 行)
- 统一初始化结构
- 保持向后兼容性
- 全面的代码格式化
**平台兼容性**
- Windows winreg API 修复(v0.52
- 安全模式匹配(无 `unwrap()`
- 跨平台托盘处理
### 配置
**同步机制**
- 跨所有应用的 MCP 同步
- Gemini 表单-编辑器同步
- 双文件读取(.env + settings.json
**验证增强**
- 输入边界检查
- TOML 引号规范化(CJK
- 自定义字段保留
- 增强的错误消息
### 代码质量
**类型安全**
- 完整的 TypeScript 覆盖
- Rust 类型改进
- API 契约验证
**测试**
- 简化的断言
- 更好的测试覆盖
- 集成测试更新
**依赖项**
- Tauri 2.8.x
- Rust`anyhow``zip``serde_yaml``tempfile`
- 前端:CodeMirror 6 包
- winreg 0.52Windows
---
## 技术统计
```
总体变更:
- 提交数:85
- 文件数:152 个文件变更
- 新增:+18,104 行
- 删除:-3,732 行
新增模块:
- Skills 管理:2,034 行(21 个文件)
- Prompts 管理:1,302 行(20 个文件)
- Gemini 集成:约 1,000 行
- MCP 重构:约 3,000 行重构
代码分布:
- 后端(Rust):约 4,500 行新增
- 前端(React):约 3,000 行新增
- 配置:约 1,500 行重构
- 测试:约 500 行
```
---
## 战略定位
### 从工具到平台
v3.7.0 代表了 CC Switch 定位的转变:
| 方面 | v3.6 | v3.7.0 |
| -------- | -------------- | ----------------------- |
| **身份** | 供应商切换器 | AI CLI 管理平台 |
| **范围** | 配置管理 | 生态系统管理 |
| **应用** | Claude + Codex | Claude + Codex + Gemini |
| **能力** | 切换配置 | 扩展能力(Skills |
| **定制** | 手动编辑 | 可视化管理(Prompts |
| **集成** | 孤立应用 | 统一管理(MCP) |
### AI CLI 管理六大支柱
1. **配置管理** - 供应商切换和管理
2. **能力扩展** - Skills 安装和生命周期
3. **行为定制** - 系统提示词预设
4. **生态集成** - 深度链接和共享
5. **多 AI 支持** - Claude/Codex/Gemini
6. **智能检测** - 冲突预防
---
## 下载与安装
### 系统要求
- **Windows**Windows 10+
- **macOS**macOS 10.15Catalina+
- **Linux**Ubuntu 22.04+ / Debian 11+ / Fedora 34+
### 下载链接
访问 [Releases](https://github.com/farion1231/cc-switch/releases/latest) 下载:
- **Windows**`CC-Switch-v3.7.0-Windows.msi``-Portable.zip`
- **macOS**`CC-Switch-v3.7.0-macOS.tar.gz``.zip`
- **Linux**`CC-Switch-v3.7.0-Linux.AppImage``.deb`
### HomebrewmacOS
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
更新:
```bash
brew upgrade --cask cc-switch
```
---
## 迁移说明
### 从 v3.6.x 升级
**自动迁移** - 无需任何操作,配置完全兼容
### 从 v3.1.x 或更早版本升级
**需要两步迁移**
1. 首先升级到 v3.2.x(执行一次性迁移)
2. 然后升级到 v3.7.0
### 新功能
- **Skills**:无需迁移,全新开始
- **Prompts**:首次启动时从 live 文件自动导入
- **Gemini**:需要单独安装 Gemini CLI
- **MCP v3.7.0**:与之前的配置向后兼容
---
## 致谢
### 贡献者
感谢所有让这个版本成为可能的贡献者:
- [@YoVinchen](https://github.com/YoVinchen) - Skills & Prompts & Geimini 集成实现
- [@farion1231](https://github.com/farion1231) - 从开发沦为 issue 回复机
- 社区成员的测试和反馈
### 赞助商
**Z.ai** - GLM CODING PLAN 赞助商
[通过此链接获得 10% 折扣](https://z.ai/subscribe?ic=8JVLJQFSKB)
**PackyCode** - API 中继服务合作伙伴
[使用 "cc-switch" 代码注册可享受 10% 折扣](https://www.packyapi.com/register?aff=cc-switch)
---
## 反馈与支持
- **问题反馈**[GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **讨论**[GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **文档**[README](../README_ZH.md)
- **更新日志**[CHANGELOG.md](../CHANGELOG.md)
---
## 未来展望
**v3.8.0 预览**(暂定):
- 本地代理功能
敬请期待更多更新!
+481
View File
@@ -0,0 +1,481 @@
# CC Switch v3.7.1
> Stability Enhancements and User Experience Improvements
**[中文更新说明 Chinese Documentation →](release-note-v3.7.1-zh.md)**
---
## v3.7.1 Updates
**Release Date**: 2025-11-22
**Code Changes**: 17 files, +524 / -81 lines
### Bug Fixes
- **Fix Third-Party Skills Installation Failure** (#268)
Fixed installation issues for skills repositories with custom subdirectories, now supports repos like `ComposioHQ/awesome-claude-skills` with subdirectories
- **Fix Gemini Configuration Persistence Issue**
Resolved the issue where settings.json edits in Gemini form were lost when switching providers
- **Prevent Dialogs from Closing on Overlay Click**
Added protection against clicking overlay/backdrop, preventing accidental form data loss across all 11 dialog components
### New Features
- **Gemini Configuration Directory Support** (#255)
Added Gemini configuration directory option in settings, supports customizing `~/.gemini/` path
- **ArchLinux Installation Support** (#259)
Added AUR installation method: `paru -S cc-switch-bin`
### Improvements
- **Skills Error Message i18n Enhancement**
Added 28+ detailed error messages (English & Chinese) with specific resolution suggestions, extended download timeout from 15s to 60s
- **Code Formatting**
Applied unified Rust and TypeScript code formatting standards
### Download
Visit [Releases](https://github.com/farion1231/cc-switch/releases/latest) to download the latest version
---
## v3.7.0 Complete Release Notes
> From Provider Switcher to All-in-One AI CLI Management Platform
**Release Date**: 2025-11-19
**Commits**: 85 from v3.6.0
**Code Changes**: 152 files, +18,104 / -3,732 lines
---
## New Features
### Gemini CLI Integration
Complete support for Google Gemini CLI, becoming the third supported application (Claude Code, Codex, Gemini).
**Core Capabilities**:
- **Dual-file configuration** - Support for both `.env` and `settings.json` formats
- **Auto-detection** - Automatically detect `GOOGLE_GEMINI_BASE_URL`, `GEMINI_MODEL`, etc.
- **Full MCP support** - Complete MCP server management for Gemini
- **Deep link integration** - Import via `ccswitch://` protocol
- **System tray** - Quick-switch from tray menu
**Provider Presets**:
- **Google Official** - OAuth authentication support
- **PackyCode** - Partner integration
- **Custom** - Full customization support
**Technical Implementation**:
- New backend modules: `gemini_config.rs` (20KB), `gemini_mcp.rs`
- Form synchronization with environment editor
- Dual-file atomic writes
---
### MCP v3.7.0 Unified Architecture
Complete refactoring of MCP management system for cross-application unification.
**Architecture Improvements**:
- **Unified panel** - Single interface for Claude/Codex/Gemini MCP servers
- **SSE transport** - New Server-Sent Events support
- **Smart parser** - Fault-tolerant JSON parsing
- **Format correction** - Auto-fix Codex `[mcp_servers]` format
- **Extended fields** - Preserve custom TOML fields
**User Experience**:
- Default app selection in forms
- JSON formatter for validation
- Improved visual hierarchy
- Better error messages
**Import/Export**:
- Unified import from all three apps
- Bidirectional synchronization
- State preservation
---
### Claude Skills Management System
**Approximately 2,000 lines of code** - A complete skill ecosystem platform.
**GitHub Integration**:
- Auto-scan skills from GitHub repositories
- Pre-configured repos:
- `ComposioHQ/awesome-claude-skills` - Curated collection
- `anthropics/skills` - Official Anthropic skills
- `cexll/myclaude` - Community contributions
- Add custom repositories
- Subdirectory scanning support (`skillsPath`)
**Lifecycle Management**:
- **Discover** - Auto-detect `SKILL.md` files
- **Install** - One-click to `~/.claude/skills/`
- **Uninstall** - Safe removal with tracking
- **Update** - Check for updates (infrastructure ready)
**Technical Architecture**:
- **Backend**: `SkillService` (526 lines) with GitHub API integration
- **Frontend**: SkillsPage, SkillCard, RepoManager
- **UI Components**: Badge, Card, Table (shadcn/ui)
- **State**: Persistent storage in `config.json`
- **i18n**: 47+ translation keys
---
### Prompts Management System
**Approximately 1,300 lines of code** - Complete system prompt management.
**Multi-Preset Management**:
- Create unlimited prompt presets
- Quick switch between presets
- One active prompt at a time
- Delete protection for active prompts
**Cross-App Support**:
- **Claude**: `~/.claude/CLAUDE.md`
- **Codex**: `~/.codex/AGENTS.md`
- **Gemini**: `~/.gemini/GEMINI.md`
**Markdown Editor**:
- Full-featured CodeMirror 6 integration
- Syntax highlighting
- Dark theme (One Dark)
- Real-time preview
**Smart Synchronization**:
- **Auto-write** - Immediately write to live files
- **Backfill protection** - Save current content before switching
- **Auto-import** - Import from live files on first launch
- **Modification protection** - Preserve manual modifications
**Technical Implementation**:
- **Backend**: `PromptService` (213 lines)
- **Frontend**: PromptPanel (177), PromptFormModal (160), MarkdownEditor (159)
- **Hooks**: usePromptActions (152 lines)
- **i18n**: 41+ translation keys
---
### Deep Link Protocol (ccswitch://)
One-click provider configuration import via URL scheme.
**Features**:
- Protocol registration on all platforms
- Import from shared links
- Lifecycle integration
- Security validation
---
### Environment Variable Conflict Detection
Intelligent detection and management of configuration conflicts.
**Detection Scope**:
- **Claude & Codex** - Cross-app conflicts
- **Gemini** - Auto-discovery
- **MCP** - Server configuration conflicts
**Management Features**:
- Visual conflict indicators
- Resolution suggestions
- Override warnings
- Backup before changes
---
## Improvements
### Provider Management
**New Presets**:
- **DouBaoSeed** - ByteDance's DouBao
- **Kimi For Coding** - Moonshot AI
- **BaiLing** - BaiLing AI
- **Removed AnyRouter** - To avoid confusion
**Enhancements**:
- Model name configuration for Codex and Gemini
- Provider notes field for organization
- Enhanced preset metadata
### Configuration Management
- **Common config migration** - From localStorage to `config.json`
- **Unified persistence** - Shared across all apps
- **Auto-import** - First launch configuration import
- **Backfill priority** - Correct handling of live files
### UI/UX Improvements
**Design System**:
- **macOS native** - System-aligned color scheme
- **Window centering** - Default centered position
- **Visual polish** - Improved spacing and hierarchy
**Interactions**:
- **Password input** - Fixed Edge/IE reveal buttons
- **URL overflow** - Fixed card overflow
- **Error copying** - Copy-to-clipboard errors
- **Tray sync** - Real-time drag-and-drop sync
---
## Bug Fixes
### Critical Fixes
- **Usage script validation** - Boundary checks
- **Gemini validation** - Relaxed constraints
- **TOML parsing** - CJK quote handling
- **MCP fields** - Custom field preservation
- **White screen** - FormLabel crash fix
### Stability
- **Tray safety** - Pattern matching instead of unwrap
- **Error isolation** - Tray failures don't block operations
- **Import classification** - Correct category assignment
### UI Fixes
- **Model placeholders** - Removed misleading hints
- **Base URL** - Auto-fill for third-party providers
- **Drag sort** - Tray menu synchronization
---
## Technical Improvements
### Architecture
**MCP v3.7.0**:
- Removed legacy code (~1,000 lines)
- Unified initialization structure
- Backward compatibility maintained
- Comprehensive code formatting
**Platform Compatibility**:
- Windows winreg API fix (v0.52)
- Safe pattern matching (no `unwrap()`)
- Cross-platform tray handling
### Configuration
**Synchronization**:
- MCP sync across all apps
- Gemini form-editor sync
- Dual-file reading (.env + settings.json)
**Validation**:
- Input boundary checks
- TOML quote normalization (CJK)
- Custom field preservation
- Enhanced error messages
### Code Quality
**Type Safety**:
- Complete TypeScript coverage
- Rust type refinements
- API contract validation
**Testing**:
- Simplified assertions
- Better test coverage
- Integration test updates
**Dependencies**:
- Tauri 2.8.x
- Rust: `anyhow`, `zip`, `serde_yaml`, `tempfile`
- Frontend: CodeMirror 6 packages
- winreg 0.52 (Windows)
---
## Technical Statistics
```
Total Changes:
- Commits: 85
- Files: 152 changed
- Additions: +18,104 lines
- Deletions: -3,732 lines
New Modules:
- Skills Management: 2,034 lines (21 files)
- Prompts Management: 1,302 lines (20 files)
- Gemini Integration: ~1,000 lines
- MCP Refactor: ~3,000 lines refactored
Code Distribution:
- Backend (Rust): ~4,500 lines new
- Frontend (React): ~3,000 lines new
- Configuration: ~1,500 lines refactored
- Tests: ~500 lines
```
---
## Strategic Positioning
### From Tool to Platform
v3.7.0 represents a shift in CC Switch's positioning:
| Aspect | v3.6 | v3.7.0 |
| ----------------- | ------------------------ | ---------------------------- |
| **Identity** | Provider Switcher | AI CLI Management Platform |
| **Scope** | Configuration Management | Ecosystem Management |
| **Applications** | Claude + Codex | Claude + Codex + Gemini |
| **Capabilities** | Switch configs | Extend capabilities (Skills) |
| **Customization** | Manual editing | Visual management (Prompts) |
| **Integration** | Isolated apps | Unified management (MCP) |
### Six Pillars of AI CLI Management
1. **Configuration Management** - Provider switching and management
2. **Capability Extension** - Skills installation and lifecycle
3. **Behavior Customization** - System prompt presets
4. **Ecosystem Integration** - Deep links and sharing
5. **Multi-AI Support** - Claude/Codex/Gemini
6. **Intelligent Detection** - Conflict prevention
---
## Download & Installation
### System Requirements
- **Windows**: Windows 10+
- **macOS**: macOS 10.15 (Catalina)+
- **Linux**: Ubuntu 22.04+ / Debian 11+ / Fedora 34+ / ArchLinux
### Download Links
Visit [Releases](https://github.com/farion1231/cc-switch/releases/latest) to download:
- **Windows**: `CC-Switch-Windows.msi` or `-Portable.zip`
- **macOS**: `CC-Switch-macOS.tar.gz` or `.zip`
- **Linux**: `CC-Switch-Linux.AppImage` or `.deb`
- **ArchLinux**: `paru -S cc-switch-bin`
### Homebrew (macOS)
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
Update:
```bash
brew upgrade --cask cc-switch
```
---
## Migration Notes
### From v3.6.x
**Automatic migration** - No action required, configs are fully compatible
### From v3.1.x or Earlier
**Two-step migration required**:
1. First upgrade to v3.2.x (performs one-time migration)
2. Then upgrade to v3.7.0
### New Features
- **Skills**: No migration needed, start fresh
- **Prompts**: Auto-import from live files on first launch
- **Gemini**: Install Gemini CLI separately if needed
- **MCP v3.7.0**: Backward compatible with previous configs
---
## Acknowledgments
### Contributors
Thanks to all contributors who made this release possible:
- [@YoVinchen](https://github.com/YoVinchen) - Skills & Prompts & Gemini integration implementation
- [@farion1231](https://github.com/farion1231) - From developer to issue responder
- Community members for testing and feedback
### Sponsors
**Z.ai** - GLM CODING PLAN sponsor
[Get 10% OFF with this link](https://z.ai/subscribe?ic=8JVLJQFSKB)
**PackyCode** - API relay service partner
[Register with "cc-switch" code for 10% discount](https://www.packyapi.com/register?aff=cc-switch)
**ShanDianShuo** - Local-first AI voice input
[Free download](https://shandianshuo.cn) for Mac/Win
---
## Feedback & Support
- **Issues**: [GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **Discussions**: [GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **Documentation**: [README](../README.md)
- **Changelog**: [CHANGELOG.md](../CHANGELOG.md)
---
## What's Next
**v3.8.0 Preview** (Tentative):
- Local proxy functionality
Stay tuned for more updates!
---
**Happy Coding!**
+481
View File
@@ -0,0 +1,481 @@
# CC Switch v3.7.1
> 稳定性增强与用户体验改进
**[English Version →](release-note-v3.7.1-en.md)**
---
## v3.7.1 更新内容
**发布日期**2025-11-22
**代码变更**17 个文件,+524 / -81 行
### Bug 修复
- **修复 Skills 第三方仓库安装失败** (#268)
修复使用自定义子目录的 skills 仓库无法安装的问题,支持类似 `ComposioHQ/awesome-claude-skills` 这样带子目录的仓库
- **修复 Gemini 配置持久化问题**
解决在 Gemini 表单中编辑 settings.json 后,切换供应商时修改丢失的问题
- **防止对话框意外关闭**
添加点击遮罩时的保护,避免误操作导致表单数据丢失,影响所有 11 个对话框组件
### 新增功能
- **Gemini 配置目录支持** (#255)
在设置中添加 Gemini 配置目录选项,支持自定义 `~/.gemini/` 路径
- **ArchLinux 安装支持** (#259)
添加 AUR 安装方式:`paru -S cc-switch-bin`
### 改进
- **Skills 错误消息国际化增强**
新增 28+ 条详细错误消息(中英文),提供具体的解决建议,下载超时从 15 秒延长到 60 秒
- **代码格式化**
应用统一的 Rust 和 TypeScript 代码格式化标准
### 下载
访问 [Releases](https://github.com/farion1231/cc-switch/releases/latest) 下载最新版本
---
## v3.7.0 完整更新说明
> 从供应商切换器到 AI CLI 一体化管理平台
**发布日期**2025-11-19
**提交数量**:从 v3.6.0 开始 85 个提交
**代码变更**152 个文件,+18,104 / -3,732 行
---
## 新增功能
### Gemini CLI 集成
完整支持 Google Gemini CLI,成为第三个支持的应用(Claude Code、Codex、Gemini)。
**核心能力**
- **双文件配置** - 同时支持 `.env``settings.json` 格式
- **自动检测** - 自动检测 `GOOGLE_GEMINI_BASE_URL``GEMINI_MODEL` 等环境变量
- **完整 MCP 支持** - 为 Gemini 提供完整的 MCP 服务器管理
- **深度链接集成** - 通过 `ccswitch://` 协议导入配置
- **系统托盘** - 从托盘菜单快速切换
**供应商预设**
- **Google Official** - 支持 OAuth 认证
- **PackyCode** - 合作伙伴集成
- **自定义** - 完全自定义支持
**技术实现**
- 新增后端模块:`gemini_config.rs`20KB)、`gemini_mcp.rs`
- 表单与环境编辑器同步
- 双文件原子写入
---
### MCP v3.7.0 统一架构
MCP 管理系统完整重构,实现跨应用统一管理。
**架构改进**
- **统一管理面板** - 单一界面管理 Claude/Codex/Gemini MCP 服务器
- **SSE 传输类型** - 新增 Server-Sent Events 支持
- **智能解析器** - 容错性 JSON 解析
- **格式修正** - 自动修复 Codex `[mcp_servers]` 格式
- **扩展字段** - 保留自定义 TOML 字段
**用户体验**
- 表单中的默认应用选择
- JSON 格式化器用于验证
- 改进的视觉层次
- 更好的错误消息
**导入/导出**
- 统一从三个应用导入
- 双向同步
- 状态保持
---
### Claude Skills 管理系统
**约 2,000 行代码** - 完整的技能生态平台。
**GitHub 集成**
- 从 GitHub 仓库自动扫描技能
- 预配置仓库:
- `ComposioHQ/awesome-claude-skills` - 精选集合
- `anthropics/skills` - Anthropic 官方技能
- `cexll/myclaude` - 社区贡献
- 添加自定义仓库
- 子目录扫描支持(`skillsPath`
**生命周期管理**
- **发现** - 自动检测 `SKILL.md` 文件
- **安装** - 一键安装到 `~/.claude/skills/`
- **卸载** - 安全移除并跟踪状态
- **更新** - 检查更新(基础设施已就绪)
**技术架构**
- **后端**`SkillService`526 行)集成 GitHub API
- **前端**SkillsPage、SkillCard、RepoManager
- **UI 组件**Badge、Card、Tableshadcn/ui
- **状态**:持久化存储在 `config.json`
- **国际化**47+ 个翻译键
---
### Prompts 管理系统
**约 1,300 行代码** - 完整的系统提示词管理。
**多预设管理**
- 创建无限数量的提示词预设
- 快速在预设间切换
- 同时只能激活一个提示词
- 活动提示词删除保护
**跨应用支持**
- **Claude**`~/.claude/CLAUDE.md`
- **Codex**`~/.codex/AGENTS.md`
- **Gemini**`~/.gemini/GEMINI.md`
**Markdown 编辑器**
- 完整的 CodeMirror 6 集成
- 语法高亮
- 暗色主题(One Dark
- 实时预览
**智能同步**
- **自动写入** - 立即写入 live 文件
- **回填保护** - 切换前保存当前内容
- **自动导入** - 首次启动从 live 文件导入
- **修改保护** - 保留手动修改
**技术实现**
- **后端**`PromptService`213 行)
- **前端**PromptPanel177)、PromptFormModal160)、MarkdownEditor159
- **Hooks**usePromptActions152 行)
- **国际化**41+ 个翻译键
---
### 深度链接协议(ccswitch://
通过 URL 方案一键导入供应商配置。
**功能特性**
- 所有平台的协议注册
- 从共享链接导入
- 生命周期集成
- 安全验证
---
### 环境变量冲突检测
智能检测和管理配置冲突。
**检测范围**
- **Claude & Codex** - 跨应用冲突
- **Gemini** - 自动发现
- **MCP** - 服务器配置冲突
**管理功能**
- 可视化冲突指示器
- 解决建议
- 覆盖警告
- 更改前备份
---
## 改进优化
### 供应商管理
**新增预设**
- **DouBaoSeed** - 字节跳动的豆包
- **Kimi For Coding** - 月之暗面
- **BaiLing** - 百灵 AI
- **移除 AnyRouter** - 避免误导
**增强功能**
- Codex 和 Gemini 的模型名称配置
- 供应商备注字段用于组织
- 增强的预设元数据
### 配置管理
- **通用配置迁移** - 从 localStorage 迁移到 `config.json`
- **统一持久化** - 跨所有应用共享
- **自动导入** - 首次启动配置导入
- **回填优先级** - 正确处理 live 文件
### UI/UX 改进
**设计系统**
- **macOS 原生** - 与系统对齐的配色方案
- **窗口居中** - 默认居中位置
- **视觉优化** - 改进的间距和层次
**交互优化**
- **密码输入** - 修复 Edge/IE 显示按钮
- **URL 溢出** - 修复卡片溢出
- **错误复制** - 可复制到剪贴板的错误
- **托盘同步** - 实时拖放同步
---
## Bug 修复
### 关键修复
- **用量脚本验证** - 边界检查
- **Gemini 验证** - 放宽约束
- **TOML 解析** - CJK 引号处理
- **MCP 字段** - 自定义字段保留
- **白屏** - FormLabel 崩溃修复
### 稳定性
- **托盘安全** - 模式匹配替代 unwrap
- **错误隔离** - 托盘失败不阻塞操作
- **导入分类** - 正确的类别分配
### UI 修复
- **模型占位符** - 移除误导性提示
- **Base URL** - 第三方供应商自动填充
- **拖拽排序** - 托盘菜单同步
---
## 技术改进
### 架构
**MCP v3.7.0**
- 移除遗留代码(约 1,000 行)
- 统一初始化结构
- 保持向后兼容性
- 全面的代码格式化
**平台兼容性**
- Windows winreg API 修复(v0.52
- 安全模式匹配(无 `unwrap()`
- 跨平台托盘处理
### 配置
**同步机制**
- 跨所有应用的 MCP 同步
- Gemini 表单-编辑器同步
- 双文件读取(.env + settings.json
**验证增强**
- 输入边界检查
- TOML 引号规范化(CJK
- 自定义字段保留
- 增强的错误消息
### 代码质量
**类型安全**
- 完整的 TypeScript 覆盖
- Rust 类型改进
- API 契约验证
**测试**
- 简化的断言
- 更好的测试覆盖
- 集成测试更新
**依赖项**
- Tauri 2.8.x
- Rust`anyhow``zip``serde_yaml``tempfile`
- 前端:CodeMirror 6 包
- winreg 0.52Windows
---
## 技术统计
```
总体变更:
- 提交数:85
- 文件数:152 个文件变更
- 新增:+18,104 行
- 删除:-3,732 行
新增模块:
- Skills 管理:2,034 行(21 个文件)
- Prompts 管理:1,302 行(20 个文件)
- Gemini 集成:约 1,000 行
- MCP 重构:约 3,000 行重构
代码分布:
- 后端(Rust):约 4,500 行新增
- 前端(React):约 3,000 行新增
- 配置:约 1,500 行重构
- 测试:约 500 行
```
---
## 战略定位
### 从工具到平台
v3.7.0 代表了 CC Switch 定位的转变:
| 方面 | v3.6 | v3.7.0 |
| -------- | -------------- | ----------------------- |
| **身份** | 供应商切换器 | AI CLI 管理平台 |
| **范围** | 配置管理 | 生态系统管理 |
| **应用** | Claude + Codex | Claude + Codex + Gemini |
| **能力** | 切换配置 | 扩展能力(Skills |
| **定制** | 手动编辑 | 可视化管理(Prompts |
| **集成** | 孤立应用 | 统一管理(MCP) |
### AI CLI 管理六大支柱
1. **配置管理** - 供应商切换和管理
2. **能力扩展** - Skills 安装和生命周期
3. **行为定制** - 系统提示词预设
4. **生态集成** - 深度链接和共享
5. **多 AI 支持** - Claude/Codex/Gemini
6. **智能检测** - 冲突预防
---
## 下载与安装
### 系统要求
- **Windows**Windows 10+
- **macOS**macOS 10.15Catalina+
- **Linux**Ubuntu 22.04+ / Debian 11+ / Fedora 34+ / ArchLinux
### 下载链接
访问 [Releases](https://github.com/farion1231/cc-switch/releases/latest) 下载:
- **Windows**`CC-Switch-Windows.msi``-Portable.zip`
- **macOS**`CC-Switch-macOS.tar.gz``.zip`
- **Linux**`CC-Switch-Linux.AppImage``.deb`
- **ArchLinux**`paru -S cc-switch-bin`
### HomebrewmacOS
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
更新:
```bash
brew upgrade --cask cc-switch
```
---
## 迁移说明
### 从 v3.6.x 升级
**自动迁移** - 无需任何操作,配置完全兼容
### 从 v3.1.x 或更早版本升级
**需要两步迁移**
1. 首先升级到 v3.2.x(执行一次性迁移)
2. 然后升级到 v3.7.0
### 新功能
- **Skills**:无需迁移,全新开始
- **Prompts**:首次启动时从 live 文件自动导入
- **Gemini**:需要单独安装 Gemini CLI
- **MCP v3.7.0**:与之前的配置向后兼容
---
## 致谢
### 贡献者
感谢所有让这个版本成为可能的贡献者:
- [@YoVinchen](https://github.com/YoVinchen) - Skills & Prompts & Gemini 集成实现
- [@farion1231](https://github.com/farion1231) - 从开发沦为 issue 回复机
- 社区成员的测试和反馈
### 赞助商
**智谱AI** - GLM CODING PLAN 赞助商
[使用此链接购买可享九折优惠](https://www.bigmodel.cn/claude-code?ic=RRVJPB5SII)
**PackyCode** - API 中转服务合作伙伴
[使用 "cc-switch" 优惠码注册享 9 折优惠](https://www.packyapi.com/register?aff=cc-switch)
**闪电说** - 本地优先的 AI 语音输入法
[免费下载](https://shandianshuo.cn) Mac/Win 双平台
---
## 反馈与支持
- **问题反馈**[GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **讨论**[GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **文档**[README](../README_ZH.md)
- **更新日志**[CHANGELOG.md](../CHANGELOG.md)
---
## 未来展望
**v3.8.0 预览**(暂定):
- 本地代理功能
敬请期待更多更新!
---
**Happy Coding!**
+369
View File
@@ -0,0 +1,369 @@
# CC Switch v3.8.0
> Persistence Architecture Upgrade, Laying the Foundation for Cloud Sync
**[中文版 →](release-note-v3.8.0-zh.md) | [日本語版 →](release-note-v3.8.0-ja.md)**
---
## Overview
CC Switch v3.8.0 is a major architectural upgrade that restructures the data persistence layer and user interface, laying the foundation for future cloud sync and local proxy features.
**Release Date**: 2025-11-28
**Commits**: 51 commits since v3.7.1
**Code Changes**: 207 files, +17,297 / -6,870 lines
---
## Major Updates
### Persistence Architecture Upgrade
Migrated from single JSON file storage to SQLite + JSON dual-layer architecture for hierarchical data management.
**Architecture Changes**:
```
v3.7.x (Old) v3.8.0 (New)
┌─────────────────┐ ┌─────────────────────────────────┐
│ config.json │ │ SQLite (Syncable Data) │
│ ┌───────────┐ │ │ ├─ providers Provider cfg │
│ │ providers │ │ │ ├─ mcp_servers MCP servers │
│ │ mcp │ │ ──> │ ├─ prompts Prompts │
│ │ prompts │ │ │ ├─ skills Skills │
│ │ settings │ │ │ └─ settings General cfg │
│ └───────────┘ │ ├─────────────────────────────────┤
└─────────────────┘ │ JSON (Device-level Data) │
│ └─ settings.json Local settings│
│ ├─ Window position │
│ ├─ Path overrides │
│ └─ Current provider ID │
└─────────────────────────────────┘
```
**Dual-layer Structure Design**:
| Layer | Storage | Data Types | Sync Strategy |
| ---------- | ------- | ------------------------------- | --------------- |
| Cloud Sync | SQLite | Providers, MCP, Prompts, Skills | Future syncable |
| Device | JSON | Window state, local paths | Stays local |
**Technical Implementation**:
- **Schema Version Management** - Supports database structure upgrade migrations
- **SQL Import/Export** - `backup.rs` supports SQL dump for cloud storage
- **Transaction Support** - SQLite native transactions ensure data consistency
- **Auto Migration** - Automatically migrates from `config.json` on first launch
**Modular Refactoring**:
```
database/
├── mod.rs Core Database struct and initialization
├── schema.rs Table definitions, schema version migrations
├── backup.rs SQL import/export, binary snapshot backup
├── migration.rs JSON → SQLite data migration engine
└── dao/ Data Access Object layer
├── providers.rs Provider CRUD
├── mcp.rs MCP server CRUD
├── prompts.rs Prompts CRUD
├── skills.rs Skills CRUD
└── settings.rs Key-value settings storage
```
---
### Brand New User Interface
Completely redesigned UI providing a more modern visual experience.
**Visual Improvements**:
- Redesigned interface layout
- Unified component styles
- Smoother transition animations
- Optimized visual hierarchy
**Interaction Enhancements**:
- Redesigned header toolbar
- Unified ConfirmDialog styling
- Disabled overscroll bounce effect on main view
- Improved form validation feedback
**Compatibility Adjustments**:
- Downgraded Tailwind CSS from v4 to v3.4 for better browser compatibility
---
### Japanese Language Support
Added Japanese interface support, expanding internationalization to three languages.
**Supported Languages**:
- Simplified Chinese
- English
- Japanese (New)
---
## New Features
### Skills Recursive Scanning
Skills management system now supports recursive scanning of repository directories, automatically discovering nested skill files.
**Improvements**:
- Support for multi-level directory structures
- Automatic discovery of all `SKILL.md` files
- Allow same-named skills from different repositories (using full path for deduplication)
---
### Provider Icon Configuration
Provider presets now support custom icon configuration.
**Features**:
- Preset providers include default icons
- Icon settings preserved when duplicating providers
- Custom icon colors
---
### Enhanced Form Validation
Provider forms now include required field validation with friendlier error messages.
**Improvements**:
- Real-time validation for required fields
- Unified Toast notifications for validation errors
- Clearer error messages
---
### Auto Launch on Startup
Added auto-launch functionality supporting Windows, macOS, and Linux platforms.
**Features**:
- One-click enable/disable in settings
- Implemented using platform-native APIs
- Windows uses Registry, macOS uses LaunchAgent, Linux uses XDG autostart
---
### New Provider Presets
- **MiniMax** - Official partner
---
## Bug Fixes
### Critical Fixes
**Custom Endpoints Lost Issue**
Fixed an issue where custom request URLs were unexpectedly lost when updating providers.
- Root Cause: `INSERT OR REPLACE` executes `DELETE + INSERT` under the hood in SQLite, triggering foreign key cascade deletion
- Fix: Changed to use `UPDATE` statement for existing providers
**Gemini Configuration Issues**
- Fixed custom provider environment variables not correctly written to `.env` file
- Fixed security auth config incorrectly written to other config files
**Provider Validation Issues**
- Fixed validation error when current provider ID doesn't exist
- Fixed icon fields lost when duplicating providers
### Platform Compatibility
**Linux**
- Resolved WebKitGTK DMA-BUF rendering issue
- Preserve user `.desktop` file customizations
### Other Fixes
- Fixed redundant usage queries when switching apps
- Fixed DMXAPI preset using wrong auth token field
- Fixed missing translation keys in deeplink components
- Fixed usage script template initialization logic
---
## Technical Improvements
### Architecture Refactoring
**Provider Service Modularization**:
```
services/provider/
├── mod.rs Core service - add/update/delete/switch/validate
├── live.rs Live config file operations
├── gemini_auth.rs Gemini auth type detection
├── endpoints.rs Custom endpoint management
└── usage.rs Usage script execution
```
**Deeplink Modularization**:
```
deeplink/
├── mod.rs Module exports
├── parser.rs URL parsing
├── provider.rs Provider import logic
├── mcp.rs MCP import logic
├── prompt.rs Prompt import
├── skill.rs Skills import
└── utils.rs Utility functions
```
### Code Quality
**Cleanup**:
- Removed legacy JSON-era import/export dead code
- Removed unused MCP type exports
- Unified error handling approach
**Test Updates**:
- Migrated tests to SQLite database architecture
- Updated component tests to match current implementation
- Fixed MSW handlers to adapt to new API
---
## Technical Statistics
```
Overall Changes:
- Commits: 51
- Files: 207 files changed
- Additions: +17,297 lines
- Deletions: -6,870 lines
- Net: +10,427 lines
Commit Type Distribution:
- fix: 25 (Bug fixes)
- refactor: 11 (Code refactoring)
- feat: 9 (New features)
- test: 1 (Testing)
- other: 5
Change Area Distribution:
- Frontend source: 112 files
- Rust backend: 63 files
- Test files: 20 files
- i18n files: 3 files
```
---
## Migration Guide
### Upgrading from v3.7.x
**Auto Migration** - Executes automatically on first launch:
1. Detects if `config.json` exists
2. Migrates all data to SQLite within a transaction
3. Migrates device-level settings to `settings.json`
4. Shows migration success notification
**Data Safety**:
- Original `config.json` file is preserved (not deleted)
- Error dialog displayed on migration failure, `config.json` preserved
- Supports Dry-run mode to verify migration logic
---
## Download & Installation
### System Requirements
- **Windows**: Windows 10+
- **macOS**: macOS 10.15 (Catalina)+
- **Linux**: Ubuntu 22.04+ / Debian 11+ / Fedora 34+
### Download Links
Visit [Releases](https://github.com/farion1231/cc-switch/releases/latest) to download:
- **Windows**: `CC-Switch-v3.8.0-Windows.msi` or `-Portable.zip`
- **macOS**: `CC-Switch-v3.8.0-macOS.tar.gz` or `.zip`
- **Linux**: `CC-Switch-v3.8.0-Linux.AppImage` or `.deb`
### Homebrew (macOS)
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
Update:
```bash
brew upgrade --cask cc-switch
```
## Acknowledgments
### Contributors
Thanks to all contributors who made this release possible:
- [@YoVinchen](https://github.com/YoVinchen) - UI and database refactoring
- [@farion1231](https://github.com/farion1231) - Bug fixes and feature enhancements
- Community members for testing and feedback
### Sponsors
**Zhipu AI** - GLM CODING PLAN Sponsor
[Get 10% off with this link](https://z.ai/subscribe?ic=8JVLJQFSKB)
**PackyCode** - API Relay Service Partner
[Use code "cc-switch" for 10% off registration](https://www.packyapi.com/register?aff=cc-switch)
**ShandianShuo** - Local-first AI Voice Input
[Free download](https://shandianshuo.cn) for Mac/Windows
**MiniMax** - MiniMax M2 CODING PLAN Sponsor
[Black Friday sale, plans starting at $2](https://platform.minimax.io/subscribe/coding-plan)
---
## Feedback & Support
- **Issue Reports**: [GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **Discussions**: [GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **Documentation**: [README](../README.md)
- **Changelog**: [CHANGELOG.md](../CHANGELOG.md)
---
## Future Roadmap
**v3.9.0 Preview** (Tentative):
- Local proxy feature
Stay tuned for more updates!
---
**Happy Coding!**
+315
View File
@@ -0,0 +1,315 @@
# CC Switch v3.8.0
> 永続化アーキテクチャを刷新し、クラウド同期の土台を構築
**[English →](release-note-v3.8.0-en.md) | [中文版 →](release-note-v3.8.0-zh.md)**
---
## 概要
CC Switch v3.8.0 はデータ永続化レイヤーと UI を大幅に作り替え、今後のクラウド同期やローカルプロキシ機能に向けた基盤を整えたメジャーアップデートです。
**リリース日**: 2025-11-28
**コミット数**: v3.7.1 以降 51 commits
**変更量**: 207 files, +17,297 / -6,870 lines
---
## 主要アップデート
### 永続化アーキテクチャの刷新
単一の JSON 保存から、階層化された SQLite + JSON の二層構造へ移行。
**アーキテクチャ変更**:
```
v3.7.x (旧) v3.8.0 (新)
┌─────────────────┐ ┌─────────────────────────────────┐
│ config.json │ │ SQLite (同期対象データ) │
│ ┌───────────┐ │ │ ├─ providers プロバイダ設定 │
│ │ providers │ │ │ ├─ mcp_servers MCP サーバー │
│ │ mcp │ │ ──> │ ├─ prompts プロンプト │
│ │ prompts │ │ │ ├─ skills Skills │
│ │ settings │ │ │ └─ settings 汎用設定 │
│ └───────────┘ │ ├─────────────────────────────────┤
└─────────────────┘ │ JSON (デバイス固有データ) │
│ └─ settings.json ローカル設定 │
│ ├─ ウィンドウ位置 │
│ ├─ パスの上書き │
│ └─ 現在のプロバイダ ID │
└─────────────────────────────────┘
```
**二層構造の設計**:
| レイヤー | ストレージ | データ種別 | 同期戦略 |
| -------- | ---------- | ----------------------------------- | ---------------- |
| クラウド | SQLite | Providers, MCP, Prompts, Skills | 将来同期対象 |
| デバイス | JSON | ウィンドウ状態、ローカルパス | ローカル保持 |
**実装ポイント**:
- **スキーマバージョン管理**: DB 構造のマイグレーションに対応
- **SQL インポート/エクスポート**: `backup.rs` が SQL ダンプをサポート
- **トランザクション対応**: SQLite ネイティブで整合性を確保
- **自動マイグレーション**: 初回起動で `config.json` から自動移行
**モジュール分割**:
```
database/
├── mod.rs Database 構造体と初期化
├── schema.rs テーブル定義とスキーマ移行
├── backup.rs SQL インポート/エクスポートとスナップショット
├── migration.rs JSON → SQLite 変換エンジン
└── dao/ DAO レイヤー
├── providers.rs プロバイダ CRUD
├── mcp.rs MCP CRUD
├── prompts.rs プロンプト CRUD
├── skills.rs Skills CRUD
└── settings.rs 設定 Key-Value 保存
```
---
### 新しいユーザーインターフェース
よりモダンな見た目と操作感に再設計。
- レイアウト全面刷新、コンポーネントスタイルを統一
- トランジションを滑らかにし、視覚的階層を最適化
- メインビューのオーバースクロールバウンスを無効化
- ブラウザ互換性向上のため Tailwind CSS を v4→v3.4 にダウングレード
---
### 日語化
UI が日本語に対応し、国際化が 3 言語(中/英/日)へ拡大。
---
## 新機能
### Skills 递帰スキャン
Skills 管理がリポジトリを再帰的に走査し、ネストされた `SKILL.md` を自動検出。
- 複数階層のディレクトリに対応
- すべての `SKILL.md` を自動発見
- パスをキーにした重複排除で同名スキルを許容
### プロバイダアイコン設定
プリセットがデフォルトアイコンを持ち、複製してもアイコンを保持。カスタム色も設定可能。
### フォームバリデーション強化
必須項目にリアルタイム検証と分かりやすいエラーメッセージを追加し、トースト通知を統一。
### 自動起動
Windows/macOS/Linux で自動起動をサポート。
- 設定画面からワンクリックで ON/OFF
- Registry / LaunchAgent / XDG autostart を使用
### 新プロバイダプリセット
- **MiniMax** - 公式パートナー
---
## バグ修正
### 重要修正
**カスタムエンドポイント消失**
- 原因: SQLite の `INSERT OR REPLACE` が内部で `DELETE + INSERT` を実行し、外部キーのカスケード削除が発生
- 対応: 既存プロバイダ更新を `UPDATE` に変更
**Gemini 設定**
- カスタム環境変数が `.env` に正しく書き込まれない問題を修正
- 認証設定が他ファイルに誤って書き込まれる問題を修正
**プロバイダ検証**
- 現在プロバイダ ID が欠落している場合のバリデーションエラーを修正
- 複製時にアイコンフィールドが失われる問題を修正
### プラットフォーム互換性
**Linux**
- WebKitGTK の DMA-BUF 描画問題を解消
- ユーザーの `.desktop` カスタマイズを保持
### その他修正
- アプリ切り替え時の不要な使用量クエリを削減
- DMXAPI プリセットの誤ったトークンフィールドを修正
- Deeplink コンポーネントの欠損翻訳キーを補完
- 使用量スクリプトテンプレート初期化を修正
---
## 技術的改善
### アーキテクチャ再編
**Provider Service のモジュール化**:
```
services/provider/
├── mod.rs 追加/更新/削除/切替/検証の中核
├── live.rs ライブ設定ファイル操作
├── gemini_auth.rs Gemini 認証タイプ検出
├── endpoints.rs カスタムエンドポイント管理
└── usage.rs 使用量スクリプト実行
```
**Deeplink のモジュール化**:
```
deeplink/
├── mod.rs エクスポート
├── parser.rs URL パース
├── provider.rs プロバイダ取り込み
├── mcp.rs MCP 取り込み
├── prompt.rs プロンプト取り込み
├── skill.rs Skills 取り込み
└── utils.rs ユーティリティ
```
### コード品質
- レガシーな JSON 時代のインポート/エクスポート死蔵コードを削除
- 使われていない MCP 型を削除し、エラーハンドリングを統一
- テストを SQLite バックエンドに移行し、MSW ハンドラを最新 API に合わせて更新
---
## 技術統計
```
全体変更:
- コミット: 51
- 変更ファイル: 207
- 追加: +17,297 行
- 削除: -6,870 行
- 純増: +10,427 行
コミット種別:
- fix: 25
- refactor: 11
- feat: 9
- test: 1
- other: 5
変更箇所:
- フロントエンド: 112 files
- Rust バックエンド: 63 files
- テスト: 20 files
- i18n: 3 files
```
---
## マイグレーションガイド
### v3.7.x からのアップグレード
**自動マイグレーション**(初回起動時):
1. `config.json` の存在を検出
2. 全データをトランザクションで SQLite に移行
3. デバイス設定を `settings.json` へ移行
4. 移行成功の通知を表示
**データ保護**:
- 元の `config.json` は保持(削除しない)
- 失敗時はエラーダイアログを表示し、`config.json` を温存
- Dry-run モードで検証可能
---
## ダウンロード & インストール
### システム要件
- **Windows**: Windows 10+
- **macOS**: macOS 10.15 (Catalina)+
- **Linux**: Ubuntu 22.04+ / Debian 11+ / Fedora 34+
### ダウンロード
[Releases](https://github.com/farion1231/cc-switch/releases/latest) から入手:
- **Windows**: `CC-Switch-v3.8.0-Windows.msi` または `-Portable.zip`
- **macOS**: `CC-Switch-v3.8.0-macOS.tar.gz` または `.zip`
- **Linux**: `CC-Switch-v3.8.0-Linux.AppImage` または `.deb`
### Homebrew (macOS)
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
アップデート:
```bash
brew upgrade --cask cc-switch
```
---
## 謝辞
### コントリビューター
- [@YoVinchen](https://github.com/YoVinchen) - UI とデータベースリファクタ
- [@farion1231](https://github.com/farion1231) - バグ修正と機能拡張
- コミュニティの皆さん - テストとフィードバック
### スポンサー
**Zhipu AI** - GLM CODING PLAN スポンサー
[10% オフリンク](https://z.ai/subscribe?ic=8JVLJQFSKB)
**PackyCode** - API リレーサービスパートナー
[登録時に「cc-switch」で 10% オフ](https://www.packyapi.com/register?aff=cc-switch)
**ShandianShuo** - ローカルファースト音声入力
[Mac/Windows 無料ダウンロード](https://shandianshuo.cn)
**MiniMax** - MiniMax M2 CODING PLAN スポンサー
[ブラックフライデーセール中、$2 から](https://platform.minimax.io/subscribe/coding-plan)
---
## フィードバック & サポート
- **Issue**: [GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **Discussions**: [GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **ドキュメント**: [README](../README.md)
- **更新履歴**: [CHANGELOG.md](../CHANGELOG.md)
---
## 今後のロードマップ
**v3.9.0 予告(予定)**:
- ローカルプロキシ機能
続報にご期待ください!
---
**Happy Coding!**
+369
View File
@@ -0,0 +1,369 @@
# CC Switch v3.8.0
> 持久化架构升级,为云同步奠定基础
**[English Version →](release-note-v3.8.0-en.md)**
---
## 概览
CC Switch v3.8.0 是一次重大的架构升级版本,重构了数据持久化层和用户界面,为未来的云同步和本地代理功能奠定基础。
**发布日期**2025-11-28
**提交数量**:从 v3.7.1 开始 51 个提交
**代码变更**207 个文件,+17,297 / -6,870 行
---
## 重大更新
### 持久化架构升级
从单一 JSON 文件存储迁移到 SQLite + JSON 双层架构,实现数据分层管理。
**架构变更**
```
v3.7.x (旧) v3.8.0 (新)
┌─────────────────┐ ┌─────────────────────────────────┐
│ config.json │ │ SQLite (可同步数据) │
│ ┌───────────┐ │ │ ├─ providers 供应商配置 │
│ │ providers │ │ │ ├─ mcp_servers MCP 服务器 │
│ │ mcp │ │ ──> │ ├─ prompts 提示词 │
│ │ prompts │ │ │ ├─ skills 技能 │
│ │ settings │ │ │ └─ settings 通用设置 │
│ └───────────┘ │ ├─────────────────────────────────┤
└─────────────────┘ │ JSON (设备级数据) │
│ └─ settings.json 本地设置 │
│ ├─ 窗口位置 │
│ ├─ 路径覆盖 │
│ └─ 当前选中供应商 ID │
└─────────────────────────────────┘
```
**双层结构设计**
| 层级 | 存储方式 | 数据类型 | 同步策略 |
| -------- | -------- | ---------------------------- | ---------- |
| 云同步层 | SQLite | 供应商、MCP、Prompts、Skills | 未来可同步 |
| 设备层 | JSON | 窗口状态、本地路径、当前选择 | 保持本地 |
**技术实现**
- **Schema 版本管理** - 支持数据库结构升级迁移
- **SQL 导入导出** - `backup.rs` 支持 SQL dump,便于云端存储
- **事务支持** - SQLite 原生事务保证数据一致性
- **自动迁移** - 首次启动自动从 `config.json` 迁移数据
**模块化重构**
```
database/
├── mod.rs 核心 Database 结构体和初始化
├── schema.rs 表结构定义、Schema 版本迁移
├── backup.rs SQL 导入导出、二进制快照备份
├── migration.rs JSON → SQLite 数据迁移引擎
└── dao/ 数据访问对象层
├── providers.rs 供应商 CRUD
├── mcp.rs MCP 服务器 CRUD
├── prompts.rs 提示词 CRUD
├── skills.rs Skills CRUD
└── settings.rs 键值对设置存储
```
---
### 全新用户界面
完整重构的 UI 设计,提供更现代化的视觉体验。
**视觉改进**
- 重新设计的界面布局
- 统一的组件样式
- 更流畅的过渡动画
- 优化的视觉层次
**交互优化**
- Header toolbar 重新设计
- ConfirmDialog 样式统一
- 禁用主视图 overscroll 弹跳效果
- 改进的表单验证反馈
**兼容性调整**
- Tailwind CSS 从 v4 降级到 v3.4,提升浏览器兼容性
---
### 日语支持
新增日语(日本語)界面支持,国际化语言扩展到三种。
**支持语言**
- 简体中文
- English
- 日本語(新增)
---
## 新增功能
### Skills 递归扫描
Skills 管理系统支持递归扫描仓库目录,自动发现嵌套的技能文件。
**改进内容**
- 支持多层目录结构
- 自动发现所有 `SKILL.md` 文件
- 允许不同仓库的同名技能(使用完整路径去重)
---
### 供应商图标配置
供应商预设支持自定义图标配置。
**功能特性**
- 预设供应商包含默认图标
- 复制供应商时保留图标设置
- 图标颜色自定义
---
### 表单验证增强
供应商表单新增必填字段验证,提供更友好的错误提示。
**改进内容**
- 必填字段实时校验
- 统一使用 Toast 通知显示验证错误
- 更清晰的错误信息
---
### 开机自启
新增开机自动启动功能,支持 Windows、macOS 和 Linux 三个平台。
**功能特性**
- 在设置中一键开启/关闭
- 使用平台原生 API 实现
- Windows 使用注册表、macOS 使用 LaunchAgent、Linux 使用 XDG autostart
---
### 新增供应商预设
- **MiniMax** - 官方合作伙伴
---
## Bug 修复
### 关键修复
**自定义端点丢失问题**
修复更新供应商时自定义请求地址意外丢失的问题。
- 根因:`INSERT OR REPLACE` 在 SQLite 底层执行 `DELETE + INSERT`,触发外键级联删除
- 修复:改用 `UPDATE` 语句更新已存在的供应商
**Gemini 配置问题**
- 修复自定义供应商环境变量未正确写入 `.env` 文件
- 修复安全认证配置错误写入到其他配置文件
**供应商验证问题**
- 修复当前供应商 ID 不存在时的验证错误
- 修复供应商复制时图标字段丢失
### 平台兼容性
**Linux**
- 解决 WebKitGTK DMA-BUF 渲染问题
- 保留用户 `.desktop` 文件自定义
### 其他修复
- 修复切换应用时的冗余用量查询
- 修复 DMXAPI 预设使用错误的认证令牌字段
- 修复深链接组件缺少翻译键
- 修复用量脚本模板初始化逻辑
---
## 技术改进
### 架构重构
**供应商服务模块化**
```
services/provider/
├── mod.rs 核心服务 - add/update/delete/switch/validate
├── live.rs Live 配置文件操作
├── gemini_auth.rs Gemini 认证类型检测
├── endpoints.rs 自定义端点管理
└── usage.rs 用量脚本执行
```
**深链接模块化**
```
deeplink/
├── mod.rs 模块导出
├── parser.rs URL 解析
├── provider.rs 供应商导入逻辑
├── mcp.rs MCP 导入逻辑
├── prompt.rs 提示词导入
├── skill.rs Skills 导入
└── utils.rs 工具函数
```
### 代码质量
**清理工作**
- 移除 JSON 时代遗留的导入导出死代码
- 移除未使用的 MCP 类型导出
- 统一错误处理方式
**测试更新**
- 迁移测试到 SQLite 数据库架构
- 更新组件测试匹配当前实现
- 修复 MSW handlers 适配新 API
---
## 技术统计
```
总体变更:
- 提交数:51
- 文件数:207 个文件变更
- 新增:+17,297 行
- 删除:-6,870 行
- 净增:+10,427 行
提交类型分布:
- fix25 个(Bug 修复)
- refactor11 个(代码重构)
- feat9 个(新功能)
- test1 个(测试)
- 其他:5 个
改动区域分布:
- 前端源码:112 个文件
- Rust 后端:63 个文件
- 测试文件:20 个文件
- 国际化文件:3 个文件
```
---
## 迁移说明
### 从 v3.7.x 升级
**自动迁移** - 首次启动时自动执行:
1. 检测 `config.json` 是否存在
2. 在事务中迁移所有数据到 SQLite
3. 设备级设置迁移到 `settings.json`
4. 显示迁移成功通知
**数据安全**
-`config.json` 文件保留不删除
- 迁移失败时显示错误对话框,保留`config.json`
- 支持 Dry-run 模式验证迁移逻辑
---
## 下载与安装
### 系统要求
- **Windows**Windows 10+
- **macOS**macOS 10.15Catalina+
- **Linux**Ubuntu 22.04+ / Debian 11+ / Fedora 34+
### 下载链接
访问 [Releases](https://github.com/farion1231/cc-switch/releases/latest) 下载:
- **Windows**`CC-Switch-v3.8.0-Windows.msi``-Portable.zip`
- **macOS**`CC-Switch-v3.8.0-macOS.tar.gz``.zip`
- **Linux**`CC-Switch-v3.8.0-Linux.AppImage``.deb`
### HomebrewmacOS
```bash
brew tap farion1231/ccswitch
brew install --cask cc-switch
```
更新:
```bash
brew upgrade --cask cc-switch
```
## 致谢
### 贡献者
感谢所有让这个版本成为可能的贡献者:
- [@YoVinchen](https://github.com/YoVinchen) - UI 和数据库重构
- [@farion1231](https://github.com/farion1231) - BUG 修复和功能增强
- 社区成员的测试和反馈
### 赞助商
**智谱AI** - GLM CODING PLAN 赞助商
[使用此链接购买可享九折优惠](https://www.bigmodel.cn/claude-code?ic=RRVJPB5SII)
**PackyCode** - API 中转服务合作伙伴
[使用 "cc-switch" 优惠码注册享 9 折优惠](https://www.packyapi.com/register?aff=cc-switch)
**闪电说** - 本地优先的 AI 语音输入法
[免费下载](https://shandianshuo.cn) Mac/Win 双平台
**MiniMax** - MiniMax M2 CODING PLAN 赞助商
[黑五优惠进行中,套餐9.9元起](https://platform.minimaxi.com/subscribe/coding-plan)
---
## 反馈与支持
- **问题反馈**[GitHub Issues](https://github.com/farion1231/cc-switch/issues)
- **讨论**[GitHub Discussions](https://github.com/farion1231/cc-switch/discussions)
- **文档**[README](../README_ZH.md)
- **更新日志**[CHANGELOG.md](../CHANGELOG.md)
---
## 未来展望
**v3.9.0 预览**(暂定):
- 本地代理功能
敬请期待更多更新!
---
**Happy Coding!**
+2 -1
View File
@@ -3,7 +3,8 @@
- mcp 管理器 ✅
- i18n ✅
- gemini cli
- homebrew 支持
- homebrew 支持
- memory 管理
- codex 更多预设供应商
- 云同步
- 本地代理
+863
View File
@@ -0,0 +1,863 @@
# v3.7.0 统一 MCP 管理重构计划
## 📋 项目概述
**目标**:将原有的按应用分离的 MCP 管理(Claude/Codex/Gemini 各自独立管理)重构为统一管理面板,每个 MCP 服务器通过多选框控制应用到哪些客户端。
**版本**v3.6.2 → v3.7.0
**开始时间**2025-11-14
---
## 🎯 核心需求
### 原有架构(v3.6.x
```
┌─────────────┐ ┌─────────────┐ ┌─────────────┐
│ Claude面板 │ │ Codex面板 │ │ Gemini面板 │
│ MCP管理 │ │ MCP管理 │ │ MCP管理 │
└─────────────┘ └─────────────┘ └─────────────┘
↓ ↓ ↓
mcp.claude mcp.codex mcp.gemini
{servers} {servers} {servers}
```
### 新架构(v3.7.0
```
┌───────────────────────────────────────┐
│ 统一 MCP 管理面板 │
│ ┌────────┬────────┬────────┬────┐ │
│ │ 服务器 │ Claude │ Codex │Gem │ │
│ ├────────┼────────┼────────┼────┤ │
│ │ mcp-1 │ ✓ │ ✓ │ │ │
│ │ mcp-2 │ ✓ │ │ ✓ │ │
│ └────────┴────────┴────────┴────┘ │
└───────────────────────────────────────┘
mcp.servers
{
"mcp-1": {
apps: {claude: true, codex: true, gemini: false}
}
}
```
---
## 📐 技术架构
### 数据结构设计
#### 新增:McpApps(应用启用状态)
```rust
#[derive(Debug, Clone, Serialize, Deserialize, Default, PartialEq)]
pub struct McpApps {
pub claude: bool,
pub codex: bool,
pub gemini: bool,
}
```
#### 更新:McpServer(统一服务器定义)
```rust
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct McpServer {
pub id: String,
pub name: String,
pub server: serde_json::Value, // 连接配置(stdio/http
pub apps: McpApps, // 新增:标记应用到哪些客户端
pub description: Option<String>,
pub homepage: Option<String>,
pub docs: Option<String>,
pub tags: Vec<String>,
}
```
#### 更新:McpRoot(新旧结构并存)
```rust
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct McpRoot {
// v3.7.0 新结构
#[serde(skip_serializing_if = "Option::is_none")]
pub servers: Option<HashMap<String, McpServer>>,
// v3.6.x 旧结构(保留用于迁移)
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub claude: McpConfig,
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub codex: McpConfig,
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub gemini: McpConfig,
}
```
### 迁移策略
```
旧配置 (v3.6.x) 新配置 (v3.7.0)
───────────────── ─────────────────
mcp: mcp:
claude: servers:
servers: mcp-fetch:
mcp-fetch: {...} → id: "mcp-fetch"
codex: server: {...}
servers: apps:
mcp-filesystem: {...} claude: true
codex: true
gemini: false
```
**迁移逻辑**
1. 检测 `mcp.servers` 是否存在
2. 若不存在,从 `mcp.claude/codex/gemini.servers` 收集所有服务器
3. 合并同 id 服务器的 apps 字段
4. 清空旧结构字段
5. 保存配置(自动触发)
---
## ✅ 开发进度
### Phase 1: 后端数据结构与迁移 ✅ 已完成
#### 1.1 修改数据结构(app_config.rs)✅
**文件**`src-tauri/src/app_config.rs`
**变更**
- ✅ 新增 `McpApps` 结构体(lines 30-62
- ✅ 新增 `McpServer` 结构体(lines 64-79
- ✅ 更新 `McpRoot` 支持新旧结构(lines 81-96
- ✅ 添加辅助方法:`is_enabled_for`, `set_enabled_for`, `enabled_apps`
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
#### 1.2 实现迁移逻辑 ✅
**文件**`src-tauri/src/app_config.rs`
**实现**
-`migrate_mcp_to_unified()` 方法(lines 380-509
- 从旧结构收集所有服务器
- 按 id 合并重复服务器
- 处理冲突(合并 apps 字段)
- 清空旧结构
- ✅ 集成到 `MultiAppConfig::load()` 方法(lines 252-257
- 自动检测并执行迁移
- 迁移后保存配置
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
---
### Phase 2: 后端服务层重构 ✅ 已完成
#### 2.1 重写 McpService ✅
**文件**`src-tauri/src/services/mcp.rs`
**新增方法**
-`get_all_servers()` - 获取所有服务器(lines 13-27
-`upsert_server()` - 添加/更新服务器(lines 30-52
-`delete_server()` - 删除服务器(lines 55-75
-`toggle_app()` - 切换应用启用状态(lines 78-111
-`sync_all_enabled()` - 同步所有启用的服务器(lines 180-188
**兼容层方法**(已废弃):
-`get_servers()` - 按应用过滤服务器(lines 196-210
-`set_enabled()` - 委托到 toggle_applines 213-222
-`sync_enabled()` - 同步特定应用(lines 225-236
-`import_from_claude/codex/gemini()` - 导入包装(lines 239-266
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
#### 2.2 新增同步函数(mcp.rs)✅
**文件**`src-tauri/src/mcp.rs`
**新增函数**lines 800-965):
-`json_server_to_toml_table()` - JSON → TOML 转换助手(lines 828-889
-`sync_single_server_to_claude()` - 同步单个服务器到 Claudelines 800-814
-`remove_server_from_claude()` - 从 Claude 移除服务器(lines 817-826
-`sync_single_server_to_codex()` - 同步单个服务器到 Codexlines 891-936
-`remove_server_from_codex()` - 从 Codex 移除服务器(lines 939-965
-`sync_single_server_to_gemini()` - 同步单个服务器到 Geminilines 967-977
-`remove_server_from_gemini()` - 从 Gemini 移除服务器(lines 980-989
**关键修复**
- ✅ 修复 toml_edit 类型转换(使用手动构建而非 serde 转换)
- ✅ 修复 get_codex_config_path() 调用(返回 PathBuf 而非 Result
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
**修复提交**`7ae2a9f` - "fix(mcp): resolve compilation errors and add backward compatibility"
#### 2.3 新增 Tauri Commands ✅
**文件**`src-tauri/src/commands/mcp.rs`
**新增命令**lines 147-196):
-`get_mcp_servers()` - 获取所有服务器(lines 154-159
-`upsert_mcp_server()` - 添加/更新服务器(lines 162-168
-`delete_mcp_server()` - 删除服务器(lines 171-177
-`toggle_mcp_app()` - 切换应用状态(lines 180-189
-`sync_all_mcp_servers()` - 同步所有服务器(lines 192-195
**更新旧命令**(兼容层):
-`upsert_mcp_server_in_config()` - 转换为统一结构(lines 68-131
-`delete_mcp_server_in_config()` - 忽略 app 参数(lines 134-141
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
**修复提交**`7ae2a9f` - "fix(mcp): resolve compilation errors and add backward compatibility"
#### 2.4 注册新命令(lib.rs)✅
**文件**`src-tauri/src/lib.rs`
**变更**
- ✅ 导出 `McpServer` 类型(line 21
- ✅ 导出新增的 mcp 同步函数(lines 26-31
- ✅ 注册 5 个新命令到 invoke_handlerlines 550-555
**提交**`c7b235b` - "feat(mcp): implement unified MCP management for v3.7.0"
#### 2.5 添加缺失的函数(claude_mcp.rs & gemini_mcp.rs)✅
**文件**
- `src-tauri/src/claude_mcp.rs` (lines 234-253)
- `src-tauri/src/gemini_mcp.rs` (lines 160-179)
**新增**
-`read_mcp_servers_map()` - 读取现有 MCP 服务器映射
**提交**`7ae2a9f` - "fix(mcp): resolve compilation errors and add backward compatibility"
#### 2.6 编译验证 ✅
**状态**:✅ 编译成功
- ⚠️ 16 个警告(8 个废弃警告 + 8 个未使用函数警告 - 预期内)
- ✅ 0 个错误
**提交**`7ae2a9f` - "fix(mcp): resolve compilation errors and add backward compatibility"
---
### Phase 3: 前端开发 ⚠️ 部分完成
#### 3.1 TypeScript 类型定义 ✅
**文件**`src/types.ts`
**变更**
- ✅ 新增 `McpApps` 接口(lines 129-133
- ✅ 更新 `McpServer` 接口(lines 136-149
- 新增 `apps: McpApps` 字段
- `name` 改为必填
- 标记 `enabled` 为废弃
- ✅ 新增 `McpServersMap` 类型别名(line 152
- ✅ 保持向后兼容(保留 `enabled`, `source` 等旧字段)
**提交**`ac09551` - "feat(frontend): add unified MCP types and API layer for v3.7.0"
#### 3.2 API 层更新 ✅
**文件**`src/lib/api/mcp.ts`
**新增方法**lines 99-141):
-`getAllServers()` - 获取所有服务器(lines 106-108
-`upsertUnifiedServer()` - 添加/更新服务器(lines 113-115
-`deleteUnifiedServer()` - 删除服务器(lines 120-122
-`toggleApp()` - 切换应用状态(lines 127-133
-`syncAllServers()` - 同步所有服务器(lines 138-140
**导入更新**
- ✅ 导入 `McpServersMap` 类型(line 6
**提交**`ac09551` - "feat(frontend): add unified MCP types and API layer for v3.7.0"
#### 3.3 React Query Hooks 📝 待开发
**计划文件**`src/hooks/useMcp.ts`
**需要实现的 Hooks**
```typescript
// 查询 hooks
export function useAllMcpServers() {
return useQuery({
queryKey: ['mcp', 'all'],
queryFn: () => mcpApi.getAllServers(),
});
}
// 变更 hooks
export function useUpsertMcpServer() {
const queryClient = useQueryClient();
return useMutation({
mutationFn: (server: McpServer) => mcpApi.upsertUnifiedServer(server),
onSuccess: () => {
queryClient.invalidateQueries({ queryKey: ['mcp', 'all'] });
},
});
}
export function useToggleMcpApp() {
const queryClient = useQueryClient();
return useMutation({
mutationFn: ({ serverId, app, enabled }: {
serverId: string;
app: AppId;
enabled: boolean;
}) => mcpApi.toggleApp(serverId, app, enabled),
onSuccess: () => {
queryClient.invalidateQueries({ queryKey: ['mcp', 'all'] });
},
});
}
export function useDeleteMcpServer() {
const queryClient = useQueryClient();
return useMutation({
mutationFn: (id: string) => mcpApi.deleteUnifiedServer(id),
onSuccess: () => {
queryClient.invalidateQueries({ queryKey: ['mcp', 'all'] });
},
});
}
export function useSyncAllMcpServers() {
return useMutation({
mutationFn: () => mcpApi.syncAllServers(),
});
}
```
**依赖**
- `@tanstack/react-query` (已安装)
- `src/lib/api/mcp.ts` (✅ 已完成)
- `src/types.ts` (✅ 已完成)
#### 3.4 统一 MCP 面板组件 📝 待开发
**计划文件**`src/components/mcp/UnifiedMcpPanel.tsx`
**组件结构**
```typescript
interface UnifiedMcpPanelProps {
className?: string;
}
export function UnifiedMcpPanel({ className }: UnifiedMcpPanelProps) {
const { t } = useTranslation();
const { data: servers, isLoading } = useAllMcpServers();
const toggleApp = useToggleMcpApp();
const deleteServer = useDeleteMcpServer();
const syncAll = useSyncAllMcpServers();
// 组件实现...
}
```
**UI 设计**
```
┌─────────────────────────────────────────────────────┐
│ MCP 服务器管理 ┌──────────┐ │
│ │ 添加服务器 │ │
│ ┌─────┐ ┌──────────────┐ ┌─────────┐ └──────────┘ │
│ │ 搜索 │ │ 导入自...▼ │ │ 同步全部 │ │
│ └─────┘ └──────────────┘ └─────────┘ │
├─────────────────────────────────────────────────────┤
│ │
│ ┌─────────────────────────────────────────────┐ │
│ │ 名称 │ Claude │ Codex │ Gemini │操作│ │
│ ├─────────────────────────────────────────────┤ │
│ │ mcp-fetch │ ✓ │ ✓ │ │ ⚙️ │ │
│ │ filesystem │ ✓ │ │ ✓ │ ⚙️ │ │
│ │ brave-search │ │ ✓ │ ✓ │ ⚙️ │ │
│ └─────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────┘
```
**功能特性**
- 📋 服务器列表展示(名称、描述、标签)
- ☑️ 三个复选框控制应用启用状态(Claude/Codex/Gemini
- 添加新服务器(表单模态框)
- ✏️ 编辑服务器(表单模态框)
- 🗑️ 删除服务器(确认对话框)
- 📥 导入功能(从 Claude/Codex/Gemini 导入)
- 🔄 同步全部(手动触发同步到 live 配置)
- 🔍 搜索过滤
- 🏷️ 标签过滤
**子组件**
1. **McpServerTable** (`McpServerTable.tsx`)
- 服务器列表表格
- 应用复选框
- 操作按钮(编辑、删除)
2. **McpServerFormModal** (`McpServerFormModal.tsx`)
- 添加/编辑表单
- stdio/http 类型切换
- 应用选择(多选)
- 元信息编辑(描述、标签、链接)
3. **McpImportDialog** (`McpImportDialog.tsx`)
- 选择导入来源(Claude/Codex/Gemini
- 服务器预览
- 批量导入
**依赖组件**(来自 shadcn/ui):
- `Table`, `TableBody`, `TableCell`, `TableHead`, `TableHeader`, `TableRow`
- `Checkbox`
- `Button`
- `Dialog`, `DialogContent`, `DialogHeader`, `DialogTitle`
- `Input`, `Textarea`, `Label`
- `Select`, `SelectContent`, `SelectItem`, `SelectTrigger`, `SelectValue`
- `Badge`
- `Tooltip`
#### 3.5 主界面集成 📝 待开发
**文件**`src/App.tsx`
**变更计划**
```typescript
// 原有代码(v3.6.x
{currentApp === 'claude' && <ClaudeMcpPanel />}
{currentApp === 'codex' && <CodexMcpPanel />}
{currentApp === 'gemini' && <GeminiMcpPanel />}
// 新代码(v3.7.0
<UnifiedMcpPanel />
```
**移除的组件**
- `ClaudeMcpPanel.tsx`
- `CodexMcpPanel.tsx`
- `GeminiMcpPanel.tsx`
**注意**:保留旧组件文件备份,以便回滚
#### 3.6 国际化文本更新 📝 待开发
**文件**
- `src/locales/zh/translation.json`
- `src/locales/en/translation.json`
**需要添加的翻译键**
```json
{
"mcp": {
"unifiedPanel": {
"title": "MCP 服务器管理 / MCP Server Management",
"addServer": "添加服务器 / Add Server",
"editServer": "编辑服务器 / Edit Server",
"deleteServer": "删除服务器 / Delete Server",
"deleteConfirm": "确定要删除此服务器吗?/ Are you sure to delete this server?",
"syncAll": "同步全部 / Sync All",
"syncAllSuccess": "已同步所有启用的服务器 / All enabled servers synced",
"importFrom": "导入自... / Import from...",
"search": "搜索服务器... / Search servers...",
"noServers": "暂无服务器 / No servers yet",
"enabledApps": "启用的应用 / Enabled Apps",
"apps": {
"claude": "Claude",
"codex": "Codex",
"gemini": "Gemini"
},
"form": {
"id": "服务器 ID / Server ID",
"name": "显示名称 / Display Name",
"type": "类型 / Type",
"stdio": "本地进程 / Local Process",
"http": "远程服务 / Remote Service",
"command": "命令 / Command",
"args": "参数 / Arguments",
"env": "环境变量 / Environment Variables",
"cwd": "工作目录 / Working Directory",
"url": "URL",
"headers": "请求头 / Headers",
"description": "描述 / Description",
"tags": "标签 / Tags",
"homepage": "主页 / Homepage",
"docs": "文档 / Documentation",
"selectApps": "选择应用 / Select Apps",
"selectAppsHint": "勾选此服务器要应用到哪些客户端 / Check which clients this server applies to"
},
"table": {
"name": "名称 / Name",
"type": "类型 / Type",
"apps": "应用 / Apps",
"actions": "操作 / Actions",
"edit": "编辑 / Edit",
"delete": "删除 / Delete"
},
"import": {
"title": "导入 MCP 服务器 / Import MCP Servers",
"fromClaude": "从 Claude 导入 / Import from Claude",
"fromCodex": "从 Codex 导入 / Import from Codex",
"fromGemini": "从 Gemini 导入 / Import from Gemini",
"success": "成功导入 {{count}} 个服务器 / Successfully imported {{count}} server(s)",
"noServersFound": "未找到可导入的服务器 / No servers found to import"
}
}
}
}
```
---
## 🔄 迁移流程
### 用户体验
```
1. 用户升级到 v3.7.0
2. 首次启动应用
3. 后端自动执行迁移
- 检测旧结构 (mcp.claude/codex/gemini.servers)
- 合并到统一结构 (mcp.servers)
- 保存迁移后的配置
- 日志记录迁移详情
4. 前端加载新面板
- 显示所有服务器
- 三个复选框显示各应用启用状态
5. 用户无缝使用
```
### 数据完整性保证
1. **迁移前验证**
- ✅ 校验旧结构合法性
- ✅ 记录迁移前状态
2. **迁移中处理**
- ✅ 合并同 id 服务器的 apps 字段
- ✅ 处理 id 冲突(保留第一个,记录警告)
- ✅ 保留所有元信息(描述、标签、链接)
3. **迁移后清理**
- ✅ 清空旧结构(claude/codex/gemini
- ✅ 自动保存新配置
- ✅ 日志记录迁移完成
4. **回滚机制**
- 配置文件有备份(`config.v1.backup.<timestamp>.json`
- 迁移失败时可手动回滚
---
## 🧪 测试计划
### 后端测试 ✅ 已验证
- [x] 编译测试(cargo check
- [x] 数据结构序列化/反序列化
- [ ] 迁移逻辑单元测试
- [ ] 服务层方法测试
- [ ] 同步函数测试
### 前端测试 ⏳ 待进行
- [ ] TypeScript 类型检查
- [ ] API 调用测试
- [ ] 组件渲染测试
- [ ] 用户交互测试
- [ ] 国际化文本检查
### 集成测试 ⏳ 待进行
- [ ] 完整迁移流程测试
- [ ] 从空配置启动
- [ ] 从 v3.6.x 配置升级
- [ ] 多服务器合并场景
- [ ] 冲突处理验证
- [ ] 多应用同步测试
- [ ] 启用单个应用
- [ ] 启用多个应用
- [ ] 动态切换应用
- [ ] 同步到 live 配置验证
- [ ] 边界情况测试
- [ ] 空服务器列表
- [ ] 超长服务器名称
- [ ] 特殊字符处理
- [ ] 并发操作
---
## 📦 交付清单
### 代码文件
#### 后端(Rust)✅ 已完成
- [x] `src-tauri/src/app_config.rs` - 数据结构定义与迁移
- [x] `src-tauri/src/services/mcp.rs` - 服务层重构
- [x] `src-tauri/src/mcp.rs` - 同步函数实现
- [x] `src-tauri/src/commands/mcp.rs` - Tauri 命令
- [x] `src-tauri/src/lib.rs` - 命令注册
- [x] `src-tauri/src/claude_mcp.rs` - Claude MCP 操作
- [x] `src-tauri/src/gemini_mcp.rs` - Gemini MCP 操作
#### 前端(TypeScript/React)⚠️ 部分完成
- [x] `src/types.ts` - 类型定义更新
- [x] `src/lib/api/mcp.ts` - API 层更新
- [ ] `src/hooks/useMcp.ts` - React Query Hooks
- [ ] `src/components/mcp/UnifiedMcpPanel.tsx` - 统一面板组件
- [ ] `src/components/mcp/McpServerTable.tsx` - 服务器表格
- [ ] `src/components/mcp/McpServerFormModal.tsx` - 表单模态框
- [ ] `src/components/mcp/McpImportDialog.tsx` - 导入对话框
- [ ] `src/App.tsx` - 主界面集成
- [ ] `src/locales/zh/translation.json` - 中文翻译
- [ ] `src/locales/en/translation.json` - 英文翻译
### 文档
- [x] 本重构计划文档 (`docs/v3.7.0-unified-mcp-refactor.md`)
- [ ] 用户升级指南 (`docs/upgrade-to-v3.7.0.md`)
- [ ] API 变更说明 (`docs/api-changes-v3.7.0.md`)
### Git 提交记录 ✅
- [x] `c7b235b` - feat(mcp): implement unified MCP management for v3.7.0
- [x] `7ae2a9f` - fix(mcp): resolve compilation errors and add backward compatibility
- [x] `ac09551` - feat(frontend): add unified MCP types and API layer for v3.7.0
---
## 🎯 下一步行动
### 立即任务(优先级 P0
1.**实现 useMcp Hook**
- 文件:`src/hooks/useMcp.ts`
- 估时:1-2 小时
- 依赖:API 层(已完成)
2.**创建 UnifiedMcpPanel 核心组件**
- 文件:`src/components/mcp/UnifiedMcpPanel.tsx`
- 估时:3-4 小时
- 依赖:useMcp Hook
3.**添加国际化文本**
- 文件:`src/locales/{zh,en}/translation.json`
- 估时:30 分钟
4.**集成到主界面**
- 文件:`src/App.tsx`
- 估时:30 分钟
- 依赖:UnifiedMcpPanel 组件
### 次要任务(优先级 P1
5.**实现子组件**
- McpServerTable
- McpServerFormModal
- McpImportDialog
- 估时:4-6 小时
6.**编写测试用例**
- 后端单元测试
- 前端组件测试
- 集成测试
- 估时:6-8 小时
7.**编写用户文档**
- 升级指南
- API 变更说明
- 估时:2-3 小时
### 优化任务(优先级 P2
8.**性能优化**
- 服务器列表虚拟滚动
- 批量操作优化
- 估时:2-3 小时
9.**用户体验增强**
- 添加加载状态
- 添加错误提示
- 添加操作确认
- 估时:2-3 小时
10.**代码清理**
- 移除旧的分应用面板组件
- 清理废弃代码
- 代码格式化
- 估时:1-2 小时
---
## 💡 技术亮点
### 1. 平滑迁移机制
- ✅ 自动检测旧配置并迁移
- ✅ 新旧结构并存(过渡期)
- ✅ 无需用户手动操作
- ✅ 保留所有历史数据
### 2. 向后兼容
- ✅ 旧命令继续可用(带废弃警告)
- ✅ 前端可增量更新
- ✅ 渐进式重构策略
### 3. 类型安全
- ✅ Rust 强类型保证数据完整性
- ✅ TypeScript 类型定义与后端一致
- ✅ serde 序列化/反序列化自动处理
### 4. 清晰的架构分层
```
Frontend (React)
↓ (Tauri IPC)
Commands Layer
Services Layer
Data Layer (Config + Live Sync)
```
### 5. SSOT 原则
- 单一配置源:`~/.cc-switch/config.json`
- 统一管理:`mcp.servers` 字段
- 按需同步:写入各应用 live 配置
---
## 📚 参考资源
### 内部文档
- [项目 README](../README.md)
- [CLAUDE.md](../CLAUDE.md) - Claude Code 工作指南
- [架构文档](../CLAUDE.md#架构概述)
### 相关 Issues/PRs
- 无(新功能开发)
### 技术栈文档
- [Tauri 2.0](https://tauri.app/v1/guides/)
- [React 18](https://react.dev/)
- [TanStack Query](https://tanstack.com/query/latest)
- [shadcn/ui](https://ui.shadcn.com/)
- [serde](https://serde.rs/)
---
## 📝 变更日志
### 2025-11-14
- ✅ 完成后端 Phase 1 & 2(数据结构、服务层、命令层)
- ✅ 修复所有编译错误
- ✅ 完成前端类型定义和 API 层
- ✅ 创建本重构计划文档
### 待更新...
---
## 👥 团队协作
**开发者**Claude Code (AI Assistant) + User
**审查者**User
**测试者**User
---
## ⚠️ 风险与对策
### 风险 1:迁移数据丢失
**概率**:低
**影响**:高
**对策**
- ✅ 迁移前自动备份配置
- ✅ 详细日志记录
- ✅ 测试各种边界情况
### 风险 2:性能问题(大量服务器)
**概率**:中
**影响**:中
**对策**
- ⬜ 实现虚拟滚动
- ⬜ 分页或懒加载
- ⬜ 性能测试
### 风险 3:兼容性问题
**概率**:中
**影响**:中
**对策**
- ✅ 保留旧命令兼容层
- ✅ 前端增量更新
- ⬜ 多版本测试
### 风险 4:用户学习成本
**概率**:低
**影响**:低
**对策**
- ⬜ 清晰的 UI 设计
- ⬜ 详细的升级指南
- ⬜ 操作提示和引导
---
## 🎉 预期收益
### 用户体验提升
-**简化操作**:不再需要在不同应用面板切换
-**统一视图**:一目了然看到所有 MCP 配置
-**灵活配置**:轻松控制每个 MCP 应用到哪些客户端
### 代码质量提升
-**架构优化**:统一数据源,消除冗余
-**维护性**:单一面板组件,代码更简洁
-**扩展性**:未来添加新应用(如 Cursor)更容易
### 性能提升
-**减少重复加载**:统一管理减少配置文件读写
-**更快同步**:批量操作更高效
---
## 📞 联系方式
**问题反馈**[GitHub Issues](https://github.com/jasonyoungyang/cc-switch/issues)
**功能建议**[GitHub Discussions](https://github.com/jasonyoungyang/cc-switch/discussions)
---
**文档版本**v1.0
**最后更新**2025-11-14
**状态**:🟡 开发中(后端完成 ✅,前端进行中 ⚠️)
+11 -5
View File
@@ -1,7 +1,7 @@
{
"name": "cc-switch",
"version": "3.6.0",
"description": "Claude Code & Codex 供应商切换工具",
"version": "3.8.2",
"description": "All-in-One Assistant for Claude Code, Codex & Gemini CLI",
"scripts": {
"dev": "pnpm tauri dev",
"build": "pnpm tauri build",
@@ -26,10 +26,13 @@
"@types/react": "^18.2.0",
"@types/react-dom": "^18.2.0",
"@vitejs/plugin-react": "^4.2.0",
"autoprefixer": "^10.4.20",
"cross-fetch": "^4.1.0",
"jsdom": "^25.0.0",
"msw": "^2.11.6",
"postcss": "^8.4.49",
"prettier": "^3.6.2",
"tailwindcss": "^3.4.17",
"typescript": "^5.3.0",
"vite": "^5.0.0",
"vitest": "^2.0.5"
@@ -37,6 +40,7 @@
"dependencies": {
"@codemirror/lang-javascript": "^6.2.4",
"@codemirror/lang-json": "^6.0.2",
"@codemirror/lang-markdown": "^6.5.0",
"@codemirror/lint": "^6.8.5",
"@codemirror/state": "^6.5.2",
"@codemirror/theme-one-dark": "^6.1.3",
@@ -45,6 +49,7 @@
"@dnd-kit/sortable": "^10.0.0",
"@dnd-kit/utilities": "^3.2.2",
"@hookform/resolvers": "^5.2.2",
"@lobehub/icons-static-svg": "^1.73.0",
"@radix-ui/react-checkbox": "^1.3.3",
"@radix-ui/react-dialog": "^1.1.15",
"@radix-ui/react-dropdown-menu": "^2.1.16",
@@ -53,7 +58,7 @@
"@radix-ui/react-slot": "^1.2.3",
"@radix-ui/react-switch": "^1.2.6",
"@radix-ui/react-tabs": "^1.1.13",
"@tailwindcss/vite": "^4.1.13",
"@radix-ui/react-visually-hidden": "^1.2.4",
"@tanstack/react-query": "^5.90.3",
"@tauri-apps/api": "^2.8.0",
"@tauri-apps/plugin-dialog": "^2.4.0",
@@ -70,10 +75,11 @@
"react-dom": "^18.2.0",
"react-hook-form": "^7.65.0",
"react-i18next": "^16.0.0",
"recharts": "^3.5.1",
"smol-toml": "^1.4.2",
"sonner": "^2.0.7",
"tailwind-merge": "^3.3.1",
"tailwindcss": "^4.1.13",
"zod": "^4.1.12"
}
},
"packageManager": "pnpm@10.10.0+sha512.d615db246fe70f25dcfea6d8d73dee782ce23e2245e3c4f6f888249fb568149318637dca73c2c5c8ef2a4ca0d5657fb9567188bfab47f566d1ee6ce987815c39"
}
+1021 -257
View File
File diff suppressed because it is too large Load Diff
+6
View File
@@ -0,0 +1,6 @@
module.exports = {
plugins: {
tailwindcss: {},
autoprefixer: {},
},
};
+208
View File
@@ -0,0 +1,208 @@
const fs = require('fs');
const path = require('path');
// 要提取的图标列表(按分类组织)
const ICONS_TO_EXTRACT = {
// AI 服务商(必需)
aiProviders: [
'openai', 'anthropic', 'claude', 'google', 'gemini',
'deepseek', 'kimi', 'moonshot', 'zhipu', 'minimax',
'baidu', 'alibaba', 'tencent', 'meta', 'microsoft',
'cohere', 'perplexity', 'mistral', 'huggingface'
],
// 云平台
cloudPlatforms: [
'aws', 'azure', 'huawei', 'cloudflare'
],
// 开发工具
devTools: [
'github', 'gitlab', 'docker', 'kubernetes', 'vscode'
],
// 其他
others: [
'settings', 'folder', 'file', 'link'
]
};
// 合并所有图标
const ALL_ICONS = [
...ICONS_TO_EXTRACT.aiProviders,
...ICONS_TO_EXTRACT.cloudPlatforms,
...ICONS_TO_EXTRACT.devTools,
...ICONS_TO_EXTRACT.others
];
// 提取逻辑
const OUTPUT_DIR = path.join(__dirname, '../src/icons/extracted');
const SOURCE_DIR = path.join(__dirname, '../node_modules/@lobehub/icons-static-svg/icons');
// 确保输出目录存在
if (!fs.existsSync(OUTPUT_DIR)) {
fs.mkdirSync(OUTPUT_DIR, { recursive: true });
}
console.log('🎨 CC-Switch Icon Extractor\n');
console.log('========================================');
console.log('📦 Extracting icons...\n');
let extracted = 0;
let notFound = [];
// 提取图标
ALL_ICONS.forEach(iconName => {
const sourceFile = path.join(SOURCE_DIR, `${iconName}.svg`);
const targetFile = path.join(OUTPUT_DIR, `${iconName}.svg`);
if (fs.existsSync(sourceFile)) {
fs.copyFileSync(sourceFile, targetFile);
console.log(`${iconName}.svg`);
extracted++;
} else {
console.log(`${iconName}.svg (not found)`);
notFound.push(iconName);
}
});
// 生成索引文件
console.log('\n📝 Generating index file...\n');
const indexContent = `// Auto-generated icon index
// Do not edit manually
export const icons: Record<string, string> = {
${ALL_ICONS.filter(name => !notFound.includes(name))
.map(name => {
const svg = fs.readFileSync(path.join(OUTPUT_DIR, `${name}.svg`), 'utf-8');
const escaped = svg.replace(/`/g, '\\`').replace(/\$/g, '\\$');
return ` '${name}': \`${escaped}\`,`;
})
.join('\n')}
};
export const iconList = Object.keys(icons);
export function getIcon(name: string): string {
return icons[name.toLowerCase()] || '';
}
export function hasIcon(name: string): boolean {
return name.toLowerCase() in icons;
}
`;
fs.writeFileSync(path.join(OUTPUT_DIR, 'index.ts'), indexContent);
console.log('✓ Generated: src/icons/extracted/index.ts');
// 生成图标元数据
const metadataContent = `// Icon metadata for search and categorization
import { IconMetadata } from '@/types/icon';
export const iconMetadata: Record<string, IconMetadata> = {
// AI Providers
openai: { name: 'openai', displayName: 'OpenAI', category: 'ai-provider', keywords: ['gpt', 'chatgpt'], defaultColor: '#00A67E' },
anthropic: { name: 'anthropic', displayName: 'Anthropic', category: 'ai-provider', keywords: ['claude'], defaultColor: '#D4915D' },
claude: { name: 'claude', displayName: 'Claude', category: 'ai-provider', keywords: ['anthropic'], defaultColor: '#D4915D' },
google: { name: 'google', displayName: 'Google', category: 'ai-provider', keywords: ['gemini', 'bard'], defaultColor: '#4285F4' },
gemini: { name: 'gemini', displayName: 'Gemini', category: 'ai-provider', keywords: ['google'], defaultColor: '#4285F4' },
deepseek: { name: 'deepseek', displayName: 'DeepSeek', category: 'ai-provider', keywords: ['deep', 'seek'], defaultColor: '#1E88E5' },
moonshot: { name: 'moonshot', displayName: 'Moonshot', category: 'ai-provider', keywords: ['kimi', 'moonshot'], defaultColor: '#6366F1' },
kimi: { name: 'kimi', displayName: 'Kimi', category: 'ai-provider', keywords: ['moonshot'], defaultColor: '#6366F1' },
zhipu: { name: 'zhipu', displayName: 'Zhipu AI', category: 'ai-provider', keywords: ['chatglm', 'glm'], defaultColor: '#0F62FE' },
minimax: { name: 'minimax', displayName: 'MiniMax', category: 'ai-provider', keywords: ['minimax'], defaultColor: '#FF6B6B' },
baidu: { name: 'baidu', displayName: 'Baidu', category: 'ai-provider', keywords: ['ernie', 'wenxin'], defaultColor: '#2932E1' },
alibaba: { name: 'alibaba', displayName: 'Alibaba', category: 'ai-provider', keywords: ['qwen', 'tongyi'], defaultColor: '#FF6A00' },
tencent: { name: 'tencent', displayName: 'Tencent', category: 'ai-provider', keywords: ['hunyuan'], defaultColor: '#00A4FF' },
meta: { name: 'meta', displayName: 'Meta', category: 'ai-provider', keywords: ['facebook', 'llama'], defaultColor: '#0081FB' },
microsoft: { name: 'microsoft', displayName: 'Microsoft', category: 'ai-provider', keywords: ['copilot', 'azure'], defaultColor: '#00A4EF' },
cohere: { name: 'cohere', displayName: 'Cohere', category: 'ai-provider', keywords: ['cohere'], defaultColor: '#39594D' },
perplexity: { name: 'perplexity', displayName: 'Perplexity', category: 'ai-provider', keywords: ['perplexity'], defaultColor: '#20808D' },
mistral: { name: 'mistral', displayName: 'Mistral', category: 'ai-provider', keywords: ['mistral'], defaultColor: '#FF7000' },
huggingface: { name: 'huggingface', displayName: 'Hugging Face', category: 'ai-provider', keywords: ['huggingface', 'hf'], defaultColor: '#FFD21E' },
// Cloud Platforms
aws: { name: 'aws', displayName: 'AWS', category: 'cloud', keywords: ['amazon', 'cloud'], defaultColor: '#FF9900' },
azure: { name: 'azure', displayName: 'Azure', category: 'cloud', keywords: ['microsoft', 'cloud'], defaultColor: '#0078D4' },
huawei: { name: 'huawei', displayName: 'Huawei', category: 'cloud', keywords: ['huawei', 'cloud'], defaultColor: '#FF0000' },
cloudflare: { name: 'cloudflare', displayName: 'Cloudflare', category: 'cloud', keywords: ['cloudflare', 'cdn'], defaultColor: '#F38020' },
// Dev Tools
github: { name: 'github', displayName: 'GitHub', category: 'tool', keywords: ['git', 'version control'], defaultColor: '#181717' },
gitlab: { name: 'gitlab', displayName: 'GitLab', category: 'tool', keywords: ['git', 'version control'], defaultColor: '#FC6D26' },
docker: { name: 'docker', displayName: 'Docker', category: 'tool', keywords: ['container'], defaultColor: '#2496ED' },
kubernetes: { name: 'kubernetes', displayName: 'Kubernetes', category: 'tool', keywords: ['k8s', 'container'], defaultColor: '#326CE5' },
vscode: { name: 'vscode', displayName: 'VS Code', category: 'tool', keywords: ['editor', 'ide'], defaultColor: '#007ACC' },
// Others
settings: { name: 'settings', displayName: 'Settings', category: 'other', keywords: ['config', 'preferences'], defaultColor: '#6B7280' },
folder: { name: 'folder', displayName: 'Folder', category: 'other', keywords: ['directory'], defaultColor: '#6B7280' },
file: { name: 'file', displayName: 'File', category: 'other', keywords: ['document'], defaultColor: '#6B7280' },
link: { name: 'link', displayName: 'Link', category: 'other', keywords: ['url', 'hyperlink'], defaultColor: '#6B7280' },
};
export function getIconMetadata(name: string): IconMetadata | undefined {
return iconMetadata[name.toLowerCase()];
}
export function searchIcons(query: string): string[] {
const lowerQuery = query.toLowerCase();
return Object.values(iconMetadata)
.filter(meta =>
meta.name.includes(lowerQuery) ||
meta.displayName.toLowerCase().includes(lowerQuery) ||
meta.keywords.some(k => k.includes(lowerQuery))
)
.map(meta => meta.name);
}
`;
fs.writeFileSync(path.join(OUTPUT_DIR, 'metadata.ts'), metadataContent);
console.log('✓ Generated: src/icons/extracted/metadata.ts');
// 生成 README
const readmeContent = `# Extracted Icons
This directory contains extracted icons from @lobehub/icons-static-svg.
## Statistics
- Total extracted: ${extracted} icons
- Not found: ${notFound.length} icons
## Extracted Icons
${ALL_ICONS.filter(name => !notFound.includes(name)).map(name => `- ${name}`).join('\n')}
${notFound.length > 0 ? `\n## Not Found\n${notFound.map(name => `- ${name}`).join('\n')}` : ''}
## Usage
\`\`\`typescript
import { getIcon, hasIcon, iconList } from './extracted';
// Get icon SVG
const svg = getIcon('openai');
// Check if icon exists
if (hasIcon('openai')) {
// ...
}
// Get all available icons
console.log(iconList);
\`\`\`
---
Last updated: ${new Date().toISOString()}
Generated by: scripts/extract-icons.js
`;
fs.writeFileSync(path.join(OUTPUT_DIR, 'README.md'), readmeContent);
console.log('✓ Generated: src/icons/extracted/README.md');
console.log('\n========================================');
console.log('✅ Extraction complete!\n');
console.log(` ✓ Extracted: ${extracted} icons`);
console.log(` ✗ Not found: ${notFound.length} icons`);
console.log(` 📉 Bundle size reduction: ~${Math.round((1 - extracted / 723) * 100)}%`);
console.log('========================================\n');
+96
View File
@@ -0,0 +1,96 @@
const fs = require('fs');
const path = require('path');
const ICONS_DIR = path.join(__dirname, '../src/icons/extracted');
// List of "Famous" icons to keep
// Based on common AI providers and tools
const KEEP_LIST = [
// AI Providers
'openai', 'anthropic', 'claude', 'google', 'gemini', 'gemma', 'palm',
'microsoft', 'azure', 'copilot', 'meta', 'llama',
'alibaba', 'qwen', 'tencent', 'hunyuan', 'baidu', 'wenxin',
'bytedance', 'doubao', 'deepseek', 'moonshot', 'kimi',
'zhipu', 'chatglm', 'glm', 'minimax', 'mistral', 'cohere',
'perplexity', 'huggingface', 'midjourney', 'stability',
'xai', 'grok', 'yi', 'zeroone', 'ollama',
'packycode',
// Cloud/Tools
'aws', 'googlecloud', 'huawei', 'cloudflare',
'github', 'githubcopilot', 'vercel', 'notion', 'discord',
'gitlab', 'docker', 'kubernetes', 'vscode', 'settings', 'folder', 'file', 'link'
];
// Get all SVG files
const files = fs.readdirSync(ICONS_DIR).filter(file => file.endsWith('.svg'));
console.log(`Scanning ${files.length} files...`);
let keptCount = 0;
let deletedCount = 0;
let renamedCount = 0;
// First pass: Identify files to keep and prefer color versions
const fileMap = {}; // name -> { hasColor: bool, hasMono: bool }
files.forEach(file => {
const isColor = file.endsWith('-color.svg');
const baseName = isColor ? file.replace('-color.svg', '') : file.replace('.svg', '');
if (!fileMap[baseName]) {
fileMap[baseName] = { hasColor: false, hasMono: false };
}
if (isColor) {
fileMap[baseName].hasColor = true;
} else {
fileMap[baseName].hasMono = true;
}
});
// Second pass: Process files
Object.keys(fileMap).forEach(baseName => {
const info = fileMap[baseName];
const shouldKeep = KEEP_LIST.includes(baseName);
if (!shouldKeep) {
// Delete both versions if not in keep list
if (info.hasColor) {
fs.unlinkSync(path.join(ICONS_DIR, `${baseName}-color.svg`));
deletedCount++;
}
if (info.hasMono) {
fs.unlinkSync(path.join(ICONS_DIR, `${baseName}.svg`));
deletedCount++;
}
return;
}
// If keeping, prefer color
if (info.hasColor) {
// Rename color version to base version (overwrite mono if exists)
const colorPath = path.join(ICONS_DIR, `${baseName}-color.svg`);
const targetPath = path.join(ICONS_DIR, `${baseName}.svg`);
try {
// If mono exists, it will be overwritten/replaced
fs.renameSync(colorPath, targetPath);
renamedCount++;
keptCount++;
} catch (e) {
console.error(`Error renaming ${baseName}:`, e);
}
} else if (info.hasMono) {
// Keep mono if no color version
keptCount++;
}
});
console.log(`\nCleanup complete:`);
console.log(`- Kept: ${keptCount}`);
console.log(`- Deleted: ${deletedCount}`);
console.log(`- Renamed (Color -> Standard): ${renamedCount}`);
// Regenerate index and metadata
require('./generate-icon-index.js');
+114
View File
@@ -0,0 +1,114 @@
const fs = require('fs');
const path = require('path');
const ICONS_DIR = path.join(__dirname, '../src/icons/extracted');
const INDEX_FILE = path.join(ICONS_DIR, 'index.ts');
const METADATA_FILE = path.join(ICONS_DIR, 'metadata.ts');
// Known metadata from previous configuration
const KNOWN_METADATA = {
openai: { name: 'openai', displayName: 'OpenAI', category: 'ai-provider', keywords: ['gpt', 'chatgpt'], defaultColor: '#00A67E' },
anthropic: { name: 'anthropic', displayName: 'Anthropic', category: 'ai-provider', keywords: ['claude'], defaultColor: '#D4915D' },
claude: { name: 'claude', displayName: 'Claude', category: 'ai-provider', keywords: ['anthropic'], defaultColor: '#D4915D' },
google: { name: 'google', displayName: 'Google', category: 'ai-provider', keywords: ['gemini', 'bard'], defaultColor: '#4285F4' },
gemini: { name: 'gemini', displayName: 'Gemini', category: 'ai-provider', keywords: ['google'], defaultColor: '#4285F4' },
deepseek: { name: 'deepseek', displayName: 'DeepSeek', category: 'ai-provider', keywords: ['deep', 'seek'], defaultColor: '#1E88E5' },
moonshot: { name: 'moonshot', displayName: 'Moonshot', category: 'ai-provider', keywords: ['kimi', 'moonshot'], defaultColor: '#6366F1' },
kimi: { name: 'kimi', displayName: 'Kimi', category: 'ai-provider', keywords: ['moonshot'], defaultColor: '#6366F1' },
zhipu: { name: 'zhipu', displayName: 'Zhipu AI', category: 'ai-provider', keywords: ['chatglm', 'glm'], defaultColor: '#0F62FE' },
minimax: { name: 'minimax', displayName: 'MiniMax', category: 'ai-provider', keywords: ['minimax'], defaultColor: '#FF6B6B' },
baidu: { name: 'baidu', displayName: 'Baidu', category: 'ai-provider', keywords: ['ernie', 'wenxin'], defaultColor: '#2932E1' },
alibaba: { name: 'alibaba', displayName: 'Alibaba', category: 'ai-provider', keywords: ['qwen', 'tongyi'], defaultColor: '#FF6A00' },
tencent: { name: 'tencent', displayName: 'Tencent', category: 'ai-provider', keywords: ['hunyuan'], defaultColor: '#00A4FF' },
meta: { name: 'meta', displayName: 'Meta', category: 'ai-provider', keywords: ['facebook', 'llama'], defaultColor: '#0081FB' },
microsoft: { name: 'microsoft', displayName: 'Microsoft', category: 'ai-provider', keywords: ['copilot', 'azure'], defaultColor: '#00A4EF' },
cohere: { name: 'cohere', displayName: 'Cohere', category: 'ai-provider', keywords: ['cohere'], defaultColor: '#39594D' },
perplexity: { name: 'perplexity', displayName: 'Perplexity', category: 'ai-provider', keywords: ['perplexity'], defaultColor: '#20808D' },
packycode: { name: 'packycode', displayName: 'PackyCode', category: 'ai-provider', keywords: ['packycode', 'packy', 'packyapi'], defaultColor: 'currentColor' },
mistral: { name: 'mistral', displayName: 'Mistral', category: 'ai-provider', keywords: ['mistral'], defaultColor: '#FF7000' },
huggingface: { name: 'huggingface', displayName: 'Hugging Face', category: 'ai-provider', keywords: ['huggingface', 'hf'], defaultColor: '#FFD21E' },
aws: { name: 'aws', displayName: 'AWS', category: 'cloud', keywords: ['amazon', 'cloud'], defaultColor: '#FF9900' },
azure: { name: 'azure', displayName: 'Azure', category: 'cloud', keywords: ['microsoft', 'cloud'], defaultColor: '#0078D4' },
huawei: { name: 'huawei', displayName: 'Huawei', category: 'cloud', keywords: ['huawei', 'cloud'], defaultColor: '#FF0000' },
cloudflare: { name: 'cloudflare', displayName: 'Cloudflare', category: 'cloud', keywords: ['cloudflare', 'cdn'], defaultColor: '#F38020' },
github: { name: 'github', displayName: 'GitHub', category: 'tool', keywords: ['git', 'version control'], defaultColor: '#181717' },
gitlab: { name: 'gitlab', displayName: 'GitLab', category: 'tool', keywords: ['git', 'version control'], defaultColor: '#FC6D26' },
docker: { name: 'docker', displayName: 'Docker', category: 'tool', keywords: ['container'], defaultColor: '#2496ED' },
kubernetes: { name: 'kubernetes', displayName: 'Kubernetes', category: 'tool', keywords: ['k8s', 'container'], defaultColor: '#326CE5' },
vscode: { name: 'vscode', displayName: 'VS Code', category: 'tool', keywords: ['editor', 'ide'], defaultColor: '#007ACC' },
settings: { name: 'settings', displayName: 'Settings', category: 'other', keywords: ['config', 'preferences'], defaultColor: '#6B7280' },
folder: { name: 'folder', displayName: 'Folder', category: 'other', keywords: ['directory'], defaultColor: '#6B7280' },
file: { name: 'file', displayName: 'File', category: 'other', keywords: ['document'], defaultColor: '#6B7280' },
link: { name: 'link', displayName: 'Link', category: 'other', keywords: ['url', 'hyperlink'], defaultColor: '#6B7280' },
};
// Get all SVG files
const files = fs.readdirSync(ICONS_DIR).filter(file => file.endsWith('.svg'));
console.log(`Found ${files.length} SVG files.`);
// Generate index.ts
const indexContent = `// Auto-generated icon index
// Do not edit manually
export const icons: Record<string, string> = {
${files.map(file => {
const name = path.basename(file, '.svg');
const svg = fs.readFileSync(path.join(ICONS_DIR, file), 'utf-8');
const escaped = svg.replace(/`/g, '\\`').replace(/\$/g, '\\$');
return ` '${name}': \`${escaped}\`,`;
}).join('\n')}
};
export const iconList = Object.keys(icons);
export function getIcon(name: string): string {
return icons[name.toLowerCase()] || '';
}
export function hasIcon(name: string): boolean {
return name.toLowerCase() in icons;
}
`;
fs.writeFileSync(INDEX_FILE, indexContent);
console.log(`Generated ${INDEX_FILE}`);
// Generate metadata.ts
const metadataEntries = files.map(file => {
const name = path.basename(file, '.svg').toLowerCase();
const known = KNOWN_METADATA[name];
if (known) {
return ` ${name}: ${JSON.stringify(known)},`;
}
// Default metadata for unknown icons
return ` '${name}': { name: '${name}', displayName: '${name}', category: 'other', keywords: [], defaultColor: 'currentColor' },`;
});
const metadataContent = `// Icon metadata for search and categorization
import { IconMetadata } from '@/types/icon';
export const iconMetadata: Record<string, IconMetadata> = {
${metadataEntries.join('\n')}
};
export function getIconMetadata(name: string): IconMetadata | undefined {
return iconMetadata[name.toLowerCase()];
}
export function searchIcons(query: string): string[] {
const lowerQuery = query.toLowerCase();
return Object.values(iconMetadata)
.filter(meta =>
meta.name.includes(lowerQuery) ||
meta.displayName.toLowerCase().includes(lowerQuery) ||
meta.keywords.some(k => k.includes(lowerQuery))
)
.map(meta => meta.name);
}
`;
fs.writeFileSync(METADATA_FILE, metadataContent);
console.log(`Generated ${METADATA_FILE}`);
+885 -14
View File
File diff suppressed because it is too large Load Diff
+33 -7
View File
@@ -1,7 +1,7 @@
[package]
name = "cc-switch"
version = "3.6.0"
description = "Claude Code & Codex 供应商配置管理工具"
version = "3.8.2"
description = "All-in-One Assistant for Claude Code, Codex & Gemini CLI"
authors = ["Jason Young"]
license = "MIT"
repository = "https://github.com/farion1231/cc-switch"
@@ -25,27 +25,49 @@ tauri-build = { version = "2.4.0", features = [] }
serde_json = "1.0"
serde = { version = "1.0", features = ["derive"] }
log = "0.4"
chrono = "0.4"
tauri = { version = "2.8.2", features = ["tray-icon"] }
chrono = { version = "0.4", features = ["serde"] }
tauri = { version = "2.8.2", features = ["tray-icon", "protocol-asset"] }
tauri-plugin-log = "2"
tauri-plugin-opener = "2"
tauri-plugin-process = "2"
tauri-plugin-updater = "2"
tauri-plugin-dialog = "2"
tauri-plugin-store = "2"
tauri-plugin-deep-link = "2"
dirs = "5.0"
toml = "0.8"
toml_edit = "0.22"
reqwest = { version = "0.12", default-features = false, features = ["rustls-tls", "json"] }
tokio = { version = "1", features = ["macros", "rt-multi-thread", "time"] }
reqwest = { version = "0.12", features = ["rustls-tls", "json", "stream"] }
tokio = { version = "1", features = ["macros", "rt-multi-thread", "time", "sync"] }
futures = "0.3"
async-stream = "0.3"
bytes = "1.5"
axum = "0.7"
tower = "0.4"
tower-http = { version = "0.5", features = ["cors"] }
hyper = { version = "1.0", features = ["full"] }
regex = "1.10"
rquickjs = { version = "0.8", features = ["array-buffer", "classes"] }
thiserror = "1.0"
thiserror = "2.0"
anyhow = "1.0"
zip = "2.2"
serde_yaml = "0.9"
tempfile = "3"
url = "2.5"
auto-launch = "0.5"
once_cell = "1.21.3"
base64 = "0.22"
rusqlite = { version = "0.31", features = ["bundled", "backup"] }
indexmap = { version = "2", features = ["serde"] }
rust_decimal = "1.33"
uuid = { version = "1.11", features = ["v4"] }
[target.'cfg(any(target_os = "macos", target_os = "windows", target_os = "linux"))'.dependencies]
tauri-plugin-single-instance = "2"
[target.'cfg(target_os = "windows")'.dependencies]
winreg = "0.52"
[target.'cfg(target_os = "macos")'.dependencies]
objc2 = "0.5"
objc2-app-kit = { version = "0.2", features = ["NSColor"] }
@@ -57,3 +79,7 @@ lto = "thin"
opt-level = "s"
panic = "abort"
strip = "symbols"
[dev-dependencies]
serial_test = "3"
tempfile = "3"
+19
View File
@@ -0,0 +1,19 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<!-- 注册 ccswitch:// 自定义 URL 协议,用于深链接导入 -->
<key>CFBundleURLTypes</key>
<array>
<dict>
<key>CFBundleURLName</key>
<string>CC Switch Deep Link</string>
<key>CFBundleURLSchemes</key>
<array>
<string>ccswitch</string>
</array>
</dict>
</array>
</dict>
</plist>
+2 -1
View File
@@ -10,7 +10,8 @@
"opener:default",
"updater:default",
"core:window:allow-set-skip-taskbar",
"core:window:allow-start-dragging",
"process:allow-restart",
"dialog:default"
]
}
}
+699 -12
View File
@@ -2,7 +2,77 @@ use serde::{Deserialize, Serialize};
use std::collections::HashMap;
use std::str::FromStr;
/// MCP 配置:单客户端维度(claude 或 codex 下的一组服务器)
use crate::services::skill::SkillStore;
/// MCP 服务器应用状态(标记应用到哪些客户端)
#[derive(Debug, Clone, Serialize, Deserialize, Default, PartialEq)]
pub struct McpApps {
#[serde(default)]
pub claude: bool,
#[serde(default)]
pub codex: bool,
#[serde(default)]
pub gemini: bool,
}
impl McpApps {
/// 检查指定应用是否启用
pub fn is_enabled_for(&self, app: &AppType) -> bool {
match app {
AppType::Claude => self.claude,
AppType::Codex => self.codex,
AppType::Gemini => self.gemini,
}
}
/// 设置指定应用的启用状态
pub fn set_enabled_for(&mut self, app: &AppType, enabled: bool) {
match app {
AppType::Claude => self.claude = enabled,
AppType::Codex => self.codex = enabled,
AppType::Gemini => self.gemini = enabled,
}
}
/// 获取所有启用的应用列表
pub fn enabled_apps(&self) -> Vec<AppType> {
let mut apps = Vec::new();
if self.claude {
apps.push(AppType::Claude);
}
if self.codex {
apps.push(AppType::Codex);
}
if self.gemini {
apps.push(AppType::Gemini);
}
apps
}
/// 检查是否所有应用都未启用
pub fn is_empty(&self) -> bool {
!self.claude && !self.codex && !self.gemini
}
}
/// MCP 服务器定义(v3.7.0 统一结构)
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct McpServer {
pub id: String,
pub name: String,
pub server: serde_json::Value,
pub apps: McpApps,
#[serde(skip_serializing_if = "Option::is_none")]
pub description: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub homepage: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub docs: Option<String>,
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub tags: Vec<String>,
}
/// MCP 配置:单客户端维度(v3.6.x 及以前,保留用于向后兼容)
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct McpConfig {
/// 以 id 为键的服务器定义(宽松 JSON 对象,包含 enabled/source 等 UI 辅助字段)
@@ -10,25 +80,72 @@ pub struct McpConfig {
pub servers: HashMap<String, serde_json::Value>,
}
/// MCP 根:按客户端分开维护(无历史兼容压力,直接以 v2 结构落地)
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
impl McpConfig {
/// 检查配置是否为空
pub fn is_empty(&self) -> bool {
self.servers.is_empty()
}
}
/// MCP 根配置(v3.7.0 新旧结构并存)
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct McpRoot {
#[serde(default)]
/// 统一的 MCP 服务器存储(v3.7.0+
#[serde(skip_serializing_if = "Option::is_none")]
pub servers: Option<HashMap<String, McpServer>>,
/// 旧的分应用存储(v3.6.x 及以前,保留用于迁移)
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub claude: McpConfig,
#[serde(default)]
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub codex: McpConfig,
#[serde(default, skip_serializing_if = "McpConfig::is_empty")]
pub gemini: McpConfig,
}
impl Default for McpRoot {
fn default() -> Self {
Self {
// v3.7.0+ 默认使用新的统一结构(空 HashMap)
servers: Some(HashMap::new()),
// 旧结构保持空,仅用于反序列化旧配置时的迁移
claude: McpConfig::default(),
codex: McpConfig::default(),
gemini: McpConfig::default(),
}
}
}
/// Prompt 配置:单客户端维度
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct PromptConfig {
#[serde(default)]
pub prompts: HashMap<String, crate::prompt::Prompt>,
}
/// Prompt 根:按客户端分开维护
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct PromptRoot {
#[serde(default)]
pub claude: PromptConfig,
#[serde(default)]
pub codex: PromptConfig,
#[serde(default)]
pub gemini: PromptConfig,
}
use crate::config::{copy_file, get_app_config_dir, get_app_config_path, write_json_file};
use crate::error::AppError;
use crate::prompt_files::prompt_file_path;
use crate::provider::ProviderManager;
/// 应用类型
#[derive(Debug, Clone, Serialize, Deserialize)]
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
#[serde(rename_all = "lowercase")]
pub enum AppType {
Claude,
Codex,
Gemini, // 新增
}
impl AppType {
@@ -36,6 +153,7 @@ impl AppType {
match self {
AppType::Claude => "claude",
AppType::Codex => "codex",
AppType::Gemini => "gemini", // 新增
}
}
}
@@ -48,15 +166,49 @@ impl FromStr for AppType {
match normalized.as_str() {
"claude" => Ok(AppType::Claude),
"codex" => Ok(AppType::Codex),
"gemini" => Ok(AppType::Gemini), // 新增
other => Err(AppError::localized(
"unsupported_app",
format!("不支持的应用标识: '{other}'。可选值: claude, codex。"),
format!("Unsupported app id: '{other}'. Allowed: claude, codex."),
format!("不支持的应用标识: '{other}'。可选值: claude, codex, gemini"),
format!("Unsupported app id: '{other}'. Allowed: claude, codex, gemini."),
)),
}
}
}
/// 通用配置片段(按应用分治)
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct CommonConfigSnippets {
#[serde(default, skip_serializing_if = "Option::is_none")]
pub claude: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub codex: Option<String>,
#[serde(default, skip_serializing_if = "Option::is_none")]
pub gemini: Option<String>,
}
impl CommonConfigSnippets {
/// 获取指定应用的通用配置片段
pub fn get(&self, app: &AppType) -> Option<&String> {
match app {
AppType::Claude => self.claude.as_ref(),
AppType::Codex => self.codex.as_ref(),
AppType::Gemini => self.gemini.as_ref(),
}
}
/// 设置指定应用的通用配置片段
pub fn set(&mut self, app: &AppType, snippet: Option<String>) {
match app {
AppType::Claude => self.claude = snippet,
AppType::Codex => self.codex = snippet,
AppType::Gemini => self.gemini = snippet,
}
}
}
/// 多应用配置结构(向后兼容)
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct MultiAppConfig {
@@ -68,6 +220,18 @@ pub struct MultiAppConfig {
/// MCP 配置(按客户端分治)
#[serde(default)]
pub mcp: McpRoot,
/// Prompt 配置(按客户端分治)
#[serde(default)]
pub prompts: PromptRoot,
/// Claude Skills 配置
#[serde(default)]
pub skills: SkillStore,
/// 通用配置片段(按应用分治)
#[serde(default)]
pub common_config_snippets: CommonConfigSnippets,
/// Claude 通用配置片段(旧字段,用于向后兼容迁移)
#[serde(default, skip_serializing_if = "Option::is_none")]
pub claude_common_config_snippet: Option<String>,
}
fn default_version() -> u32 {
@@ -79,11 +243,16 @@ impl Default for MultiAppConfig {
let mut apps = HashMap::new();
apps.insert("claude".to_string(), ProviderManager::default());
apps.insert("codex".to_string(), ProviderManager::default());
apps.insert("gemini".to_string(), ProviderManager::default()); // 新增
Self {
version: 2,
apps,
mcp: McpRoot::default(),
prompts: PromptRoot::default(),
skills: SkillStore::default(),
common_config_snippets: CommonConfigSnippets::default(),
claude_common_config_snippet: None,
}
}
}
@@ -94,8 +263,12 @@ impl MultiAppConfig {
let config_path = get_app_config_path();
if !config_path.exists() {
log::info!("配置文件不存在,创建新的多应用配置");
return Ok(Self::default());
log::info!("配置文件不存在,创建新的多应用配置并自动导入提示词");
// 使用新的方法,支持自动导入提示词
let config = Self::default_with_auto_import()?;
// 立即保存到磁盘
config.save()?;
return Ok(config);
}
// 尝试读取文件
@@ -121,8 +294,73 @@ impl MultiAppConfig {
));
}
let has_skills_in_config = value
.as_object()
.is_some_and(|map| map.contains_key("skills"));
// 解析 v2 结构
serde_json::from_value::<Self>(value).map_err(|e| AppError::json(&config_path, e))
let mut config: Self =
serde_json::from_value(value).map_err(|e| AppError::json(&config_path, e))?;
let mut updated = false;
if !has_skills_in_config {
let skills_path = get_app_config_dir().join("skills.json");
if skills_path.exists() {
match std::fs::read_to_string(&skills_path) {
Ok(content) => match serde_json::from_str::<SkillStore>(&content) {
Ok(store) => {
config.skills = store;
updated = true;
log::info!("已从旧版 skills.json 导入 Claude Skills 配置");
}
Err(e) => {
log::warn!("解析旧版 skills.json 失败: {e}");
}
},
Err(e) => {
log::warn!("读取旧版 skills.json 失败: {e}");
}
}
}
}
// 确保 gemini 应用存在(兼容旧配置文件)
if !config.apps.contains_key("gemini") {
config
.apps
.insert("gemini".to_string(), ProviderManager::default());
updated = true;
}
// 执行 MCP 迁移(v3.6.x → v3.7.0
let migrated = config.migrate_mcp_to_unified()?;
if migrated {
log::info!("MCP 配置已迁移到 v3.7.0 统一结构,保存配置...");
updated = true;
}
// 对于已经存在的配置文件,如果此前版本还没有 Prompt 功能,
// 且 prompts 仍然是空的,则尝试自动导入现有提示词文件。
let imported_prompts = config.maybe_auto_import_prompts_for_existing_config()?;
if imported_prompts {
updated = true;
}
// 迁移通用配置片段:claude_common_config_snippet → common_config_snippets.claude
if let Some(old_claude_snippet) = config.claude_common_config_snippet.take() {
log::info!(
"迁移通用配置:claude_common_config_snippet → common_config_snippets.claude"
);
config.common_config_snippets.claude = Some(old_claude_snippet);
updated = true;
}
if updated {
log::info!("配置结构已更新(包括 MCP 迁移或 Prompt 自动导入),保存配置...");
config.save()?;
}
Ok(config)
}
/// 保存配置到文件
@@ -132,7 +370,7 @@ impl MultiAppConfig {
if config_path.exists() {
let backup_path = get_app_config_dir().join("config.json.bak");
if let Err(e) = copy_file(&config_path, &backup_path) {
log::warn!("备份 config.json 到 .bak 失败: {}", e);
log::warn!("备份 config.json 到 .bak 失败: {e}");
}
}
@@ -163,6 +401,7 @@ impl MultiAppConfig {
match app {
AppType::Claude => &self.mcp.claude,
AppType::Codex => &self.mcp.codex,
AppType::Gemini => &self.mcp.gemini,
}
}
@@ -171,6 +410,454 @@ impl MultiAppConfig {
match app {
AppType::Claude => &mut self.mcp.claude,
AppType::Codex => &mut self.mcp.codex,
AppType::Gemini => &mut self.mcp.gemini,
}
}
/// 创建默认配置并自动导入已存在的提示词文件
fn default_with_auto_import() -> Result<Self, AppError> {
log::info!("首次启动,创建默认配置并检测提示词文件");
let mut config = Self::default();
// 为每个应用尝试自动导入提示词
Self::auto_import_prompt_if_exists(&mut config, AppType::Claude)?;
Self::auto_import_prompt_if_exists(&mut config, AppType::Codex)?;
Self::auto_import_prompt_if_exists(&mut config, AppType::Gemini)?;
Ok(config)
}
/// 已存在配置文件时的 Prompt 自动导入逻辑
///
/// 适用于「老版本已经生成过 config.json,但当时还没有 Prompt 功能」的升级场景。
/// 判定规则:
/// - 仅当所有应用的 prompts 都为空时才尝试导入(避免打扰已经在使用 Prompt 功能的用户)
/// - 每个应用最多导入一次,对应各自的提示词文件(如 CLAUDE.md/AGENTS.md/GEMINI.md
///
/// 返回值:
/// - Ok(true) 表示至少有一个应用成功导入了提示词
/// - Ok(false) 表示无需导入或未导入任何内容
fn maybe_auto_import_prompts_for_existing_config(&mut self) -> Result<bool, AppError> {
// 如果任一应用已经有提示词配置,说明用户已经在使用 Prompt 功能,避免再次自动导入
if !self.prompts.claude.prompts.is_empty()
|| !self.prompts.codex.prompts.is_empty()
|| !self.prompts.gemini.prompts.is_empty()
{
return Ok(false);
}
log::info!("检测到已存在配置文件且 Prompt 列表为空,将尝试从现有提示词文件自动导入");
let mut imported = false;
for app in [AppType::Claude, AppType::Codex, AppType::Gemini] {
// 复用已有的单应用导入逻辑
if Self::auto_import_prompt_if_exists(self, app)? {
imported = true;
}
}
Ok(imported)
}
/// 检查并自动导入单个应用的提示词文件
///
/// 返回值:
/// - Ok(true) 表示成功导入了非空文件
/// - Ok(false) 表示未导入(文件不存在、内容为空或读取失败)
fn auto_import_prompt_if_exists(config: &mut Self, app: AppType) -> Result<bool, AppError> {
let file_path = prompt_file_path(&app)?;
// 检查文件是否存在
if !file_path.exists() {
log::debug!("提示词文件不存在,跳过自动导入: {file_path:?}");
return Ok(false);
}
// 读取文件内容
let content = match std::fs::read_to_string(&file_path) {
Ok(c) => c,
Err(e) => {
log::warn!("读取提示词文件失败: {file_path:?}, 错误: {e}");
return Ok(false); // 失败时不中断,继续处理其他应用
}
};
// 检查内容是否为空
if content.trim().is_empty() {
log::debug!("提示词文件内容为空,跳过导入: {file_path:?}");
return Ok(false);
}
log::info!("发现提示词文件,自动导入: {file_path:?}");
// 创建提示词对象
let timestamp = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.map(|d| d.as_secs() as i64)
.unwrap_or_else(|_| {
log::warn!("Failed to get system time, using 0 as timestamp");
0
});
let id = format!("auto-imported-{timestamp}");
let prompt = crate::prompt::Prompt {
id: id.clone(),
name: format!(
"Auto-imported Prompt {}",
chrono::Local::now().format("%Y-%m-%d %H:%M")
),
content,
description: Some("Automatically imported on first launch".to_string()),
enabled: true, // 自动启用
created_at: Some(timestamp),
updated_at: Some(timestamp),
};
// 插入到对应的应用配置中
let prompts = match app {
AppType::Claude => &mut config.prompts.claude.prompts,
AppType::Codex => &mut config.prompts.codex.prompts,
AppType::Gemini => &mut config.prompts.gemini.prompts,
};
prompts.insert(id, prompt);
log::info!("自动导入完成: {}", app.as_str());
Ok(true)
}
/// 将 v3.6.x 的分应用 MCP 结构迁移到 v3.7.0 的统一结构
///
/// 迁移策略:
/// 1. 检查是否已经迁移(mcp.servers 是否存在)
/// 2. 收集所有应用的 MCP,按 ID 去重合并
/// 3. 生成统一的 McpServer 结构,标记应用到哪些客户端
/// 4. 清空旧的分应用配置
pub fn migrate_mcp_to_unified(&mut self) -> Result<bool, AppError> {
// 检查是否已经是新结构
if self.mcp.servers.is_some() {
log::debug!("MCP 配置已是统一结构,跳过迁移");
return Ok(false);
}
log::info!("检测到旧版 MCP 配置格式,开始迁移到 v3.7.0 统一结构...");
let mut unified_servers: HashMap<String, McpServer> = HashMap::new();
let mut conflicts = Vec::new();
// 收集所有应用的 MCP
for app in [AppType::Claude, AppType::Codex, AppType::Gemini] {
let old_servers = match app {
AppType::Claude => &self.mcp.claude.servers,
AppType::Codex => &self.mcp.codex.servers,
AppType::Gemini => &self.mcp.gemini.servers,
};
for (id, entry) in old_servers {
let enabled = entry
.get("enabled")
.and_then(|v| v.as_bool())
.unwrap_or(true);
if let Some(existing) = unified_servers.get_mut(id) {
// 该 ID 已存在,合并 apps 字段
existing.apps.set_enabled_for(&app, enabled);
// 检测配置冲突(同 ID 但配置不同)
if existing.server != *entry.get("server").unwrap_or(&serde_json::json!({})) {
conflicts.push(format!(
"MCP '{id}' 在 {} 和之前的应用中配置不同,将使用首次遇到的配置",
app.as_str()
));
}
} else {
// 首次遇到该 MCP,创建新条目
let name = entry
.get("name")
.and_then(|v| v.as_str())
.unwrap_or(id)
.to_string();
let server = entry
.get("server")
.cloned()
.unwrap_or(serde_json::json!({}));
let description = entry
.get("description")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
let homepage = entry
.get("homepage")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
let docs = entry
.get("docs")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
let tags = entry
.get("tags")
.and_then(|v| v.as_array())
.map(|arr| {
arr.iter()
.filter_map(|v| v.as_str().map(|s| s.to_string()))
.collect()
})
.unwrap_or_default();
let mut apps = McpApps::default();
apps.set_enabled_for(&app, enabled);
unified_servers.insert(
id.clone(),
McpServer {
id: id.clone(),
name,
server,
apps,
description,
homepage,
docs,
tags,
},
);
}
}
}
// 记录冲突警告
if !conflicts.is_empty() {
log::warn!("MCP 迁移过程中检测到配置冲突:");
for conflict in &conflicts {
log::warn!(" - {conflict}");
}
}
log::info!(
"MCP 迁移完成,共迁移 {} 个服务器{}",
unified_servers.len(),
if !conflicts.is_empty() {
format!("(存在 {} 个冲突)", conflicts.len())
} else {
String::new()
}
);
// 替换为新结构
self.mcp.servers = Some(unified_servers);
// 清空旧的分应用配置
self.mcp.claude = McpConfig::default();
self.mcp.codex = McpConfig::default();
self.mcp.gemini = McpConfig::default();
Ok(true)
}
}
#[cfg(test)]
mod tests {
use super::*;
use serial_test::serial;
use std::env;
use std::fs;
use tempfile::TempDir;
struct TempHome {
#[allow(dead_code)] // 字段通过 Drop trait 管理临时目录生命周期
dir: TempDir,
original_home: Option<String>,
original_userprofile: Option<String>,
}
impl TempHome {
fn new() -> Self {
let dir = TempDir::new().expect("failed to create temp home");
let original_home = env::var("HOME").ok();
let original_userprofile = env::var("USERPROFILE").ok();
env::set_var("HOME", dir.path());
env::set_var("USERPROFILE", dir.path());
Self {
dir,
original_home,
original_userprofile,
}
}
}
impl Drop for TempHome {
fn drop(&mut self) {
match &self.original_home {
Some(value) => env::set_var("HOME", value),
None => env::remove_var("HOME"),
}
match &self.original_userprofile {
Some(value) => env::set_var("USERPROFILE", value),
None => env::remove_var("USERPROFILE"),
}
}
}
fn write_prompt_file(app: AppType, content: &str) {
let path = crate::prompt_files::prompt_file_path(&app).expect("prompt path");
if let Some(parent) = path.parent() {
fs::create_dir_all(parent).expect("create parent dir");
}
fs::write(path, content).expect("write prompt");
}
#[test]
#[serial]
fn auto_imports_existing_prompt_when_config_missing() {
let _home = TempHome::new();
write_prompt_file(AppType::Claude, "# hello");
let config = MultiAppConfig::load().expect("load config");
assert_eq!(config.prompts.claude.prompts.len(), 1);
let prompt = config
.prompts
.claude
.prompts
.values()
.next()
.expect("prompt exists");
assert!(prompt.enabled);
assert_eq!(prompt.content, "# hello");
let config_path = crate::config::get_app_config_path();
assert!(
config_path.exists(),
"auto import should persist config to disk"
);
}
#[test]
#[serial]
fn skips_empty_prompt_files_during_import() {
let _home = TempHome::new();
write_prompt_file(AppType::Claude, " \n ");
let config = MultiAppConfig::load().expect("load config");
assert!(
config.prompts.claude.prompts.is_empty(),
"empty files must be ignored"
);
}
#[test]
#[serial]
fn auto_import_happens_only_once() {
let _home = TempHome::new();
write_prompt_file(AppType::Claude, "first version");
let first = MultiAppConfig::load().expect("load config");
assert_eq!(first.prompts.claude.prompts.len(), 1);
let claude_prompt = first
.prompts
.claude
.prompts
.values()
.next()
.expect("prompt exists")
.content
.clone();
assert_eq!(claude_prompt, "first version");
// 覆盖文件内容,但保留 config.json
write_prompt_file(AppType::Claude, "second version");
let second = MultiAppConfig::load().expect("load config again");
assert_eq!(second.prompts.claude.prompts.len(), 1);
let prompt = second
.prompts
.claude
.prompts
.values()
.next()
.expect("prompt exists");
assert_eq!(
prompt.content, "first version",
"should not re-import when config already exists"
);
}
#[test]
#[serial]
fn auto_imports_gemini_prompt_on_first_launch() {
let _home = TempHome::new();
write_prompt_file(AppType::Gemini, "# Gemini Prompt\n\nTest content");
let config = MultiAppConfig::load().expect("load config");
assert_eq!(config.prompts.gemini.prompts.len(), 1);
let prompt = config
.prompts
.gemini
.prompts
.values()
.next()
.expect("gemini prompt exists");
assert!(prompt.enabled, "gemini prompt should be enabled");
assert_eq!(prompt.content, "# Gemini Prompt\n\nTest content");
assert_eq!(
prompt.description,
Some("Automatically imported on first launch".to_string())
);
}
#[test]
#[serial]
fn auto_imports_all_three_apps_prompts() {
let _home = TempHome::new();
write_prompt_file(AppType::Claude, "# Claude prompt");
write_prompt_file(AppType::Codex, "# Codex prompt");
write_prompt_file(AppType::Gemini, "# Gemini prompt");
let config = MultiAppConfig::load().expect("load config");
// 验证所有三个应用的提示词都被导入
assert_eq!(config.prompts.claude.prompts.len(), 1);
assert_eq!(config.prompts.codex.prompts.len(), 1);
assert_eq!(config.prompts.gemini.prompts.len(), 1);
// 验证所有提示词都被启用
assert!(
config
.prompts
.claude
.prompts
.values()
.next()
.unwrap()
.enabled
);
assert!(
config
.prompts
.codex
.prompts
.values()
.next()
.unwrap()
.enabled
);
assert!(
config
.prompts
.gemini
.prompts
.values()
.next()
.unwrap()
.enabled
);
}
}
+8 -12
View File
@@ -30,7 +30,7 @@ fn read_override_from_store(app: &tauri::AppHandle) -> Option<PathBuf> {
let store = match app.store_builder("app_paths.json").build() {
Ok(store) => store,
Err(e) => {
log::warn!("无法创建 Store: {}", e);
log::warn!("无法创建 Store: {e}");
return None;
}
};
@@ -46,21 +46,17 @@ fn read_override_from_store(app: &tauri::AppHandle) -> Option<PathBuf> {
if !path.exists() {
log::warn!(
"Store 中配置的 app_config_dir 不存在: {:?}\n\
将使用默认路径。",
path
"Store 中配置的 app_config_dir 不存在: {path:?}\n\
将使用默认路径。"
);
return None;
}
log::info!("使用 Store 中的 app_config_dir: {:?}", path);
log::info!("使用 Store 中的 app_config_dir: {path:?}");
Some(path)
}
Some(_) => {
log::warn!(
"Store 中的 {} 类型不正确,应为字符串",
STORE_KEY_APP_CONFIG_DIR
);
log::warn!("Store 中的 {STORE_KEY_APP_CONFIG_DIR} 类型不正确,应为字符串");
None
}
None => None,
@@ -82,14 +78,14 @@ pub fn set_app_config_dir_to_store(
let store = app
.store_builder("app_paths.json")
.build()
.map_err(|e| AppError::Message(format!("创建 Store 失败: {}", e)))?;
.map_err(|e| AppError::Message(format!("创建 Store 失败: {e}")))?;
match path {
Some(p) => {
let trimmed = p.trim();
if !trimmed.is_empty() {
store.set(STORE_KEY_APP_CONFIG_DIR, Value::String(trimmed.to_string()));
log::info!("已将 app_config_dir 写入 Store: {}", trimmed);
log::info!("已将 app_config_dir 写入 Store: {trimmed}");
} else {
store.delete(STORE_KEY_APP_CONFIG_DIR);
log::info!("已从 Store 中删除 app_config_dir 配置");
@@ -103,7 +99,7 @@ pub fn set_app_config_dir_to_store(
store
.save()
.map_err(|e| AppError::Message(format!("保存 Store 失败: {}", e)))?;
.map_err(|e| AppError::Message(format!("保存 Store 失败: {e}")))?;
refresh_app_config_dir_override(app);
Ok(())
+49
View File
@@ -0,0 +1,49 @@
use crate::error::AppError;
use auto_launch::{AutoLaunch, AutoLaunchBuilder};
/// 初始化 AutoLaunch 实例
fn get_auto_launch() -> Result<AutoLaunch, AppError> {
let app_name = "CC Switch";
let app_path =
std::env::current_exe().map_err(|e| AppError::Message(format!("无法获取应用路径: {e}")))?;
// 使用 AutoLaunchBuilder 消除平台差异
// Windows/Linux: new() 接受 3 参数
// macOS: new() 接受 4 参数(含 hidden 参数)
// Builder 模式自动处理这些差异
let auto_launch = AutoLaunchBuilder::new()
.set_app_name(app_name)
.set_app_path(&app_path.to_string_lossy())
.build()
.map_err(|e| AppError::Message(format!("创建 AutoLaunch 失败: {e}")))?;
Ok(auto_launch)
}
/// 启用开机自启
pub fn enable_auto_launch() -> Result<(), AppError> {
let auto_launch = get_auto_launch()?;
auto_launch
.enable()
.map_err(|e| AppError::Message(format!("启用开机自启失败: {e}")))?;
log::info!("已启用开机自启");
Ok(())
}
/// 禁用开机自启
pub fn disable_auto_launch() -> Result<(), AppError> {
let auto_launch = get_auto_launch()?;
auto_launch
.disable()
.map_err(|e| AppError::Message(format!("禁用开机自启失败: {e}")))?;
log::info!("已禁用开机自启");
Ok(())
}
/// 检查是否已启用开机自启
pub fn is_auto_launch_enabled() -> Result<bool, AppError> {
let auto_launch = get_auto_launch()?;
auto_launch
.is_enabled()
.map_err(|e| AppError::Message(format!("检查开机自启状态失败: {e}")))
}
+30 -11
View File
@@ -37,7 +37,7 @@ fn ensure_mcp_override_migrated() {
if let Some(parent) = new_path.parent() {
if let Err(err) = fs::create_dir_all(parent) {
log::warn!("创建 MCP 目录失败: {}", err);
log::warn!("创建 MCP 目录失败: {err}");
return;
}
}
@@ -118,9 +118,10 @@ pub fn upsert_mcp_server(id: &str, spec: Value) -> Result<bool, AppError> {
let t_opt = spec.get("type").and_then(|x| x.as_str());
let is_stdio = t_opt.map(|t| t == "stdio").unwrap_or(true); // 兼容缺省(按 stdio 处理)
let is_http = t_opt.map(|t| t == "http").unwrap_or(false);
if !(is_stdio || is_http) {
let is_sse = t_opt.map(|t| t == "sse").unwrap_or(false);
if !(is_stdio || is_http || is_sse) {
return Err(AppError::McpValidation(
"MCP 服务器 type 必须是 'stdio''http'(或省略表示 stdio".into(),
"MCP 服务器 type 必须是 'stdio''http' 或 'sse'(或省略表示 stdio".into(),
));
}
@@ -134,13 +135,15 @@ pub fn upsert_mcp_server(id: &str, spec: Value) -> Result<bool, AppError> {
}
}
// http 类型必须有 url
if is_http {
// http/sse 类型必须有 url
if is_http || is_sse {
let url = spec.get("url").and_then(|x| x.as_str()).unwrap_or("");
if url.is_empty() {
return Err(AppError::McpValidation(
"http 类型的 MCP 服务器缺少 url 字段".into(),
));
return Err(AppError::McpValidation(if is_http {
"http 类型的 MCP 服务器缺少 url 字段".into()
} else {
"sse 类型的 MCP 服务器缺少 url 字段".into()
}));
}
}
@@ -231,6 +234,23 @@ pub fn validate_command_in_path(cmd: &str) -> Result<bool, AppError> {
Ok(false)
}
/// 读取 ~/.claude.json 中的 mcpServers 映射
pub fn read_mcp_servers_map() -> Result<std::collections::HashMap<String, Value>, AppError> {
let path = user_config_path();
if !path.exists() {
return Ok(std::collections::HashMap::new());
}
let root = read_json_value(&path)?;
let servers = root
.get("mcpServers")
.and_then(|v| v.as_object())
.map(|obj| obj.iter().map(|(k, v)| (k.clone(), v.clone())).collect())
.unwrap_or_default();
Ok(servers)
}
/// 将给定的启用 MCP 服务器映射写入到用户级 ~/.claude.json 的 mcpServers 字段
/// 仅覆盖 mcpServers,其他字段保持不变
pub fn set_mcp_servers_map(
@@ -250,14 +270,13 @@ pub fn set_mcp_servers_map(
map.clone()
} else {
return Err(AppError::McpValidation(format!(
"MCP 服务器 '{}' 不是对象",
id
"MCP 服务器 '{id}' 不是对象"
)));
};
if let Some(server_val) = obj.remove("server") {
let server_obj = server_val.as_object().cloned().ok_or_else(|| {
AppError::McpValidation(format!("MCP 服务器 '{}' server 字段不是对象", id))
AppError::McpValidation(format!("MCP 服务器 '{id}' server 字段不是对象"))
})?;
obj = server_obj;
}
+2 -2
View File
@@ -80,7 +80,7 @@ pub fn write_claude_config() -> Result<bool, AppError> {
if changed || !path.exists() {
let serialized = serde_json::to_string_pretty(&obj)
.map_err(|e| AppError::JsonSerialize { source: e })?;
fs::write(&path, format!("{}\n", serialized)).map_err(|e| AppError::io(&path, e))?;
fs::write(&path, format!("{serialized}\n")).map_err(|e| AppError::io(&path, e))?;
Ok(true)
} else {
Ok(false)
@@ -114,7 +114,7 @@ pub fn clear_claude_config() -> Result<bool, AppError> {
let serialized =
serde_json::to_string_pretty(&value).map_err(|e| AppError::JsonSerialize { source: e })?;
fs::write(&path, format!("{}\n", serialized)).map_err(|e| AppError::io(&path, e))?;
fs::write(&path, format!("{serialized}\n")).map_err(|e| AppError::io(&path, e))?;
Ok(true)
}
+4 -2
View File
@@ -29,6 +29,7 @@ pub fn get_codex_config_path() -> PathBuf {
}
/// 获取 Codex 供应商配置文件路径
#[allow(dead_code)]
pub fn get_codex_provider_paths(
provider_id: &str,
provider_name: Option<&str>,
@@ -37,13 +38,14 @@ pub fn get_codex_provider_paths(
.map(sanitize_provider_name)
.unwrap_or_else(|| sanitize_provider_name(provider_id));
let auth_path = get_codex_config_dir().join(format!("auth-{}.json", base_name));
let config_path = get_codex_config_dir().join(format!("config-{}.toml", base_name));
let auth_path = get_codex_config_dir().join(format!("auth-{base_name}.json"));
let config_path = get_codex_config_dir().join(format!("config-{base_name}.toml"));
(auth_path, config_path)
}
/// 删除 Codex 供应商配置文件
#[allow(dead_code)]
pub fn delete_codex_provider_config(
provider_id: &str,
provider_name: &str,
+101 -6
View File
@@ -29,6 +29,15 @@ pub async fn get_config_status(app: String) -> Result<ConfigStatus, String> {
Ok(ConfigStatus { exists, path })
}
AppType::Gemini => {
let env_path = crate::gemini_config::get_gemini_env_path();
let exists = env_path.exists();
let path = crate::gemini_config::get_gemini_dir()
.to_string_lossy()
.to_string();
Ok(ConfigStatus { exists, path })
}
}
}
@@ -44,6 +53,7 @@ pub async fn get_config_dir(app: String) -> Result<String, String> {
let dir = match AppType::from_str(&app).map_err(|e| e.to_string())? {
AppType::Claude => config::get_claude_config_dir(),
AppType::Codex => codex_config::get_codex_config_dir(),
AppType::Gemini => crate::gemini_config::get_gemini_dir(),
};
Ok(dir.to_string_lossy().to_string())
@@ -55,16 +65,17 @@ pub async fn open_config_folder(handle: AppHandle, app: String) -> Result<bool,
let config_dir = match AppType::from_str(&app).map_err(|e| e.to_string())? {
AppType::Claude => config::get_claude_config_dir(),
AppType::Codex => codex_config::get_codex_config_dir(),
AppType::Gemini => crate::gemini_config::get_gemini_dir(),
};
if !config_dir.exists() {
std::fs::create_dir_all(&config_dir).map_err(|e| format!("创建目录失败: {}", e))?;
std::fs::create_dir_all(&config_dir).map_err(|e| format!("创建目录失败: {e}"))?;
}
handle
.opener()
.open_path(config_dir.to_string_lossy().to_string(), None::<String>)
.map_err(|e| format!("打开文件夹失败: {}", e))?;
.map_err(|e| format!("打开文件夹失败: {e}"))?;
Ok(true)
}
@@ -87,14 +98,14 @@ pub async fn pick_directory(
builder.blocking_pick_folder()
})
.await
.map_err(|e| format!("弹出目录选择器失败: {}", e))?;
.map_err(|e| format!("弹出目录选择器失败: {e}"))?;
match result {
Some(file_path) => {
let resolved = file_path
.simplified()
.into_path()
.map_err(|e| format!("解析选择的目录失败: {}", e))?;
.map_err(|e| format!("解析选择的目录失败: {e}"))?;
Ok(Some(resolved.to_string_lossy().to_string()))
}
None => Ok(None),
@@ -114,13 +125,97 @@ pub async fn open_app_config_folder(handle: AppHandle) -> Result<bool, String> {
let config_dir = config::get_app_config_dir();
if !config_dir.exists() {
std::fs::create_dir_all(&config_dir).map_err(|e| format!("创建目录失败: {}", e))?;
std::fs::create_dir_all(&config_dir).map_err(|e| format!("创建目录失败: {e}"))?;
}
handle
.opener()
.open_path(config_dir.to_string_lossy().to_string(), None::<String>)
.map_err(|e| format!("打开文件夹失败: {}", e))?;
.map_err(|e| format!("打开文件夹失败: {e}"))?;
Ok(true)
}
/// 获取 Claude 通用配置片段(已废弃,使用 get_common_config_snippet
#[tauri::command]
pub async fn get_claude_common_config_snippet(
state: tauri::State<'_, crate::store::AppState>,
) -> Result<Option<String>, String> {
state
.db
.get_config_snippet("claude")
.map_err(|e| e.to_string())
}
/// 设置 Claude 通用配置片段(已废弃,使用 set_common_config_snippet
#[tauri::command]
pub async fn set_claude_common_config_snippet(
snippet: String,
state: tauri::State<'_, crate::store::AppState>,
) -> Result<(), String> {
// 验证是否为有效的 JSON(如果不为空)
if !snippet.trim().is_empty() {
serde_json::from_str::<serde_json::Value>(&snippet)
.map_err(|e| format!("无效的 JSON 格式: {e}"))?;
}
let value = if snippet.trim().is_empty() {
None
} else {
Some(snippet)
};
state
.db
.set_config_snippet("claude", value)
.map_err(|e| e.to_string())?;
Ok(())
}
/// 获取通用配置片段(统一接口)
#[tauri::command]
pub async fn get_common_config_snippet(
app_type: String,
state: tauri::State<'_, crate::store::AppState>,
) -> Result<Option<String>, String> {
state
.db
.get_config_snippet(&app_type)
.map_err(|e| e.to_string())
}
/// 设置通用配置片段(统一接口)
#[tauri::command]
pub async fn set_common_config_snippet(
app_type: String,
snippet: String,
state: tauri::State<'_, crate::store::AppState>,
) -> Result<(), String> {
// 验证格式(根据应用类型)
if !snippet.trim().is_empty() {
match app_type.as_str() {
"claude" | "gemini" => {
// 验证 JSON 格式
serde_json::from_str::<serde_json::Value>(&snippet)
.map_err(|e| format!("无效的 JSON 格式: {e}"))?;
}
"codex" => {
// TOML 格式暂不验证(或可使用 toml crate)
// 注意:TOML 验证较为复杂,暂时跳过
}
_ => {}
}
}
let value = if snippet.trim().is_empty() {
None
} else {
Some(snippet)
};
state
.db
.set_config_snippet(&app_type, value)
.map_err(|e| e.to_string())?;
Ok(())
}
+89
View File
@@ -0,0 +1,89 @@
use crate::deeplink::{
import_mcp_from_deeplink, import_prompt_from_deeplink, import_provider_from_deeplink,
import_skill_from_deeplink, parse_deeplink_url, DeepLinkImportRequest,
};
use crate::store::AppState;
use tauri::State;
/// Parse a deep link URL and return the parsed request for frontend confirmation
#[tauri::command]
pub fn parse_deeplink(url: String) -> Result<DeepLinkImportRequest, String> {
log::info!("Parsing deep link URL: {url}");
parse_deeplink_url(&url).map_err(|e| e.to_string())
}
/// Merge configuration from Base64/URL into a deep link request
/// This is used by the frontend to show the complete configuration in the confirmation dialog
#[tauri::command]
pub fn merge_deeplink_config(
request: DeepLinkImportRequest,
) -> Result<DeepLinkImportRequest, String> {
log::info!("Merging config for deep link request: {:?}", request.name);
crate::deeplink::parse_and_merge_config(&request).map_err(|e| e.to_string())
}
/// Import a provider from a deep link request (legacy, kept for compatibility)
#[tauri::command]
pub fn import_from_deeplink(
state: State<AppState>,
request: DeepLinkImportRequest,
) -> Result<String, String> {
log::info!(
"Importing provider from deep link: {:?} for app {:?}",
request.name,
request.app
);
let provider_id = import_provider_from_deeplink(&state, request).map_err(|e| e.to_string())?;
log::info!("Successfully imported provider with ID: {provider_id}");
Ok(provider_id)
}
/// Import resource from a deep link request (unified handler)
#[tauri::command]
pub async fn import_from_deeplink_unified(
state: State<'_, AppState>,
request: DeepLinkImportRequest,
) -> Result<serde_json::Value, String> {
log::info!("Importing {} resource from deep link", request.resource);
match request.resource.as_str() {
"provider" => {
let provider_id =
import_provider_from_deeplink(&state, request).map_err(|e| e.to_string())?;
Ok(serde_json::json!({
"type": "provider",
"id": provider_id
}))
}
"prompt" => {
let prompt_id =
import_prompt_from_deeplink(&state, request).map_err(|e| e.to_string())?;
Ok(serde_json::json!({
"type": "prompt",
"id": prompt_id
}))
}
"mcp" => {
let result = import_mcp_from_deeplink(&state, request).map_err(|e| e.to_string())?;
// Add type field to the result
Ok(serde_json::json!({
"type": "mcp",
"importedCount": result.imported_count,
"importedIds": result.imported_ids,
"failed": result.failed
}))
}
"skill" => {
let skill_key =
import_skill_from_deeplink(&state, request).map_err(|e| e.to_string())?;
Ok(serde_json::json!({
"type": "skill",
"key": skill_key
}))
}
_ => Err(format!("Unsupported resource type: {}", request.resource)),
}
}
+22
View File
@@ -0,0 +1,22 @@
use crate::services::env_checker::{check_env_conflicts as check_conflicts, EnvConflict};
use crate::services::env_manager::{
delete_env_vars as delete_vars, restore_from_backup, BackupInfo,
};
/// Check environment variable conflicts for a specific app
#[tauri::command]
pub fn check_env_conflicts(app: String) -> Result<Vec<EnvConflict>, String> {
check_conflicts(&app)
}
/// Delete environment variables with backup
#[tauri::command]
pub fn delete_env_vars(conflicts: Vec<EnvConflict>) -> Result<BackupInfo, String> {
delete_vars(conflicts)
}
/// Restore environment variables from backup file
#[tauri::command]
pub fn restore_env_backup(backup_path: String) -> Result<(), String> {
restore_from_backup(backup_path)
}
+45 -40
View File
@@ -6,73 +6,78 @@ use tauri::State;
use tauri_plugin_dialog::DialogExt;
use crate::error::AppError;
use crate::services::ConfigService;
use crate::services::provider::ProviderService;
use crate::store::AppState;
/// 导出配置文件
/// 导出数据库为 SQL 备份
#[tauri::command]
pub async fn export_config_to_file(
#[allow(non_snake_case)] filePath: String,
state: State<'_, AppState>,
) -> Result<Value, String> {
let db = state.db.clone();
tauri::async_runtime::spawn_blocking(move || {
let target_path = PathBuf::from(&filePath);
ConfigService::export_config_to_path(&target_path)?;
db.export_sql(&target_path)?;
Ok::<_, AppError>(json!({
"success": true,
"message": "Configuration exported successfully",
"message": "SQL exported successfully",
"filePath": filePath
}))
})
.await
.map_err(|e| format!("导出配置失败: {}", e))?
.map_err(|e| format!("导出配置失败: {e}"))?
.map_err(|e: AppError| e.to_string())
}
/// 从文件导入配置
/// 从 SQL 备份导入数据库
#[tauri::command]
pub async fn import_config_from_file(
#[allow(non_snake_case)] filePath: String,
state: State<'_, AppState>,
) -> Result<Value, String> {
let (new_config, backup_id) = tauri::async_runtime::spawn_blocking(move || {
let db = state.db.clone();
let db_for_state = db.clone();
tauri::async_runtime::spawn_blocking(move || {
let path_buf = PathBuf::from(&filePath);
ConfigService::load_config_for_import(&path_buf)
let backup_id = db.import_sql(&path_buf)?;
// 导入后同步当前供应商到各自的 live 配置
let app_state = AppState::new(db_for_state);
if let Err(err) = ProviderService::sync_current_to_live(&app_state) {
log::warn!("导入后同步 live 配置失败: {err}");
}
// 重新加载设置到内存缓存,确保导入的设置生效
if let Err(err) = crate::settings::reload_settings() {
log::warn!("导入后重载设置失败: {err}");
}
Ok::<_, AppError>(json!({
"success": true,
"message": "SQL imported successfully",
"backupId": backup_id
}))
})
.await
.map_err(|e| format!("导入配置失败: {}", e))?
.map_err(|e: AppError| e.to_string())?;
{
let mut guard = state
.config
.write()
.map_err(|e| AppError::from(e).to_string())?;
*guard = new_config;
}
Ok(json!({
"success": true,
"message": "Configuration imported successfully",
"backupId": backup_id
}))
.map_err(|e| format!("导入配置失败: {e}"))?
.map_err(|e: AppError| e.to_string())
}
/// 同步当前供应商配置到对应的 live 文件
#[tauri::command]
pub async fn sync_current_providers_live(state: State<'_, AppState>) -> Result<Value, String> {
{
let mut config_state = state
.config
.write()
.map_err(|e| AppError::from(e).to_string())?;
ConfigService::sync_current_providers_to_live(&mut config_state)
.map_err(|e| e.to_string())?;
}
Ok(json!({
"success": true,
"message": "Live configuration synchronized"
}))
let db = state.db.clone();
tauri::async_runtime::spawn_blocking(move || {
let app_state = AppState::new(db);
ProviderService::sync_current_to_live(&app_state)?;
Ok::<_, AppError>(json!({
"success": true,
"message": "Live configuration synchronized"
}))
})
.await
.map_err(|e| format!("同步当前供应商失败: {e}"))?
.map_err(|e: AppError| e.to_string())
}
/// 保存文件对话框
@@ -84,7 +89,7 @@ pub async fn save_file_dialog<R: tauri::Runtime>(
let dialog = app.dialog();
let result = dialog
.file()
.add_filter("JSON", &["json"])
.add_filter("SQL", &["sql"])
.set_file_name(&defaultName)
.blocking_save_file();
@@ -99,7 +104,7 @@ pub async fn open_file_dialog<R: tauri::Runtime>(
let dialog = app.dialog();
let result = dialog
.file()
.add_filter("JSON", &["json"])
.add_filter("SQL", &["sql"])
.blocking_pick_file();
Ok(result.map(|p| p.to_string()))
+83 -20
View File
@@ -1,5 +1,6 @@
#![allow(non_snake_case)]
use indexmap::IndexMap;
use std::collections::HashMap;
use serde::Serialize;
@@ -50,6 +51,7 @@ pub struct McpConfigResponse {
use std::str::FromStr;
#[tauri::command]
#[allow(deprecated)] // 兼容层命令,内部调用已废弃的 Service 方法
pub async fn get_mcp_config(
state: State<'_, AppState>,
app: String,
@@ -66,6 +68,7 @@ pub async fn get_mcp_config(
}
/// 在 config.json 中新增或更新一个 MCP 服务器定义
/// [已废弃] 该命令仍然使用旧的分应用API,会转换为统一结构
#[tauri::command]
pub async fn upsert_mcp_server_in_config(
state: State<'_, AppState>,
@@ -74,8 +77,55 @@ pub async fn upsert_mcp_server_in_config(
spec: serde_json::Value,
sync_other_side: Option<bool>,
) -> Result<bool, String> {
use crate::app_config::McpServer;
let app_ty = AppType::from_str(&app).map_err(|e| e.to_string())?;
McpService::upsert_server(&state, app_ty, &id, spec, sync_other_side.unwrap_or(false))
// 读取现有的服务器(如果存在)
let existing_server = {
let servers = state.db.get_all_mcp_servers().map_err(|e| e.to_string())?;
servers.get(&id).cloned()
};
// 构建新的统一服务器结构
let mut new_server = if let Some(mut existing) = existing_server {
// 更新现有服务器
existing.server = spec.clone();
existing.apps.set_enabled_for(&app_ty, true);
existing
} else {
// 创建新服务器
let mut apps = crate::app_config::McpApps::default();
apps.set_enabled_for(&app_ty, true);
// 尝试从 spec 中提取 name,否则使用 id
let name = spec
.get("name")
.and_then(|v| v.as_str())
.unwrap_or(&id)
.to_string();
McpServer {
id: id.clone(),
name,
server: spec,
apps,
description: None,
homepage: None,
docs: None,
tags: Vec::new(),
}
};
// 如果 sync_other_side 为 true,也启用其他应用
if sync_other_side.unwrap_or(false) {
new_server.apps.claude = true;
new_server.apps.codex = true;
new_server.apps.gemini = true;
}
McpService::upsert_server(&state, new_server)
.map(|_| true)
.map_err(|e| e.to_string())
}
@@ -83,15 +133,15 @@ pub async fn upsert_mcp_server_in_config(
#[tauri::command]
pub async fn delete_mcp_server_in_config(
state: State<'_, AppState>,
app: String,
_app: String, // 参数保留用于向后兼容,但在统一结构中不再需要
id: String,
) -> Result<bool, String> {
let app_ty = AppType::from_str(&app).map_err(|e| e.to_string())?;
McpService::delete_server(&state, app_ty, &id).map_err(|e| e.to_string())
McpService::delete_server(&state, &id).map_err(|e| e.to_string())
}
/// 设置启用状态并同步到客户端配置
#[tauri::command]
#[allow(deprecated)] // 兼容层命令,内部调用已废弃的 Service 方法
pub async fn set_mcp_enabled(
state: State<'_, AppState>,
app: String,
@@ -102,30 +152,43 @@ pub async fn set_mcp_enabled(
McpService::set_enabled(&state, app_ty, &id, enabled).map_err(|e| e.to_string())
}
/// 手动同步:将启用的 MCP 投影到 ~/.claude.json
// ============================================================================
// v3.7.0 新增:统一 MCP 管理命令
// ============================================================================
use crate::app_config::McpServer;
/// 获取所有 MCP 服务器(统一结构)
#[tauri::command]
pub async fn sync_enabled_mcp_to_claude(state: State<'_, AppState>) -> Result<bool, String> {
McpService::sync_enabled(&state, AppType::Claude)
.map(|_| true)
.map_err(|e| e.to_string())
pub async fn get_mcp_servers(
state: State<'_, AppState>,
) -> Result<IndexMap<String, McpServer>, String> {
McpService::get_all_servers(&state).map_err(|e| e.to_string())
}
/// 手动同步:将启用的 MCP 投影到 ~/.codex/config.toml
/// 添加或更新 MCP 服务器
#[tauri::command]
pub async fn sync_enabled_mcp_to_codex(state: State<'_, AppState>) -> Result<bool, String> {
McpService::sync_enabled(&state, AppType::Codex)
.map(|_| true)
.map_err(|e| e.to_string())
pub async fn upsert_mcp_server(
state: State<'_, AppState>,
server: McpServer,
) -> Result<(), String> {
McpService::upsert_server(&state, server).map_err(|e| e.to_string())
}
/// 从 ~/.claude.json 导入 MCP 定义到 config.json
/// 删除 MCP 服务器
#[tauri::command]
pub async fn import_mcp_from_claude(state: State<'_, AppState>) -> Result<usize, String> {
McpService::import_from_claude(&state).map_err(|e| e.to_string())
pub async fn delete_mcp_server(state: State<'_, AppState>, id: String) -> Result<bool, String> {
McpService::delete_server(&state, &id).map_err(|e| e.to_string())
}
/// 从 ~/.codex/config.toml 导入 MCP 定义到 config.json
/// 切换 MCP 服务器在指定应用的启用状态
#[tauri::command]
pub async fn import_mcp_from_codex(state: State<'_, AppState>) -> Result<usize, String> {
McpService::import_from_codex(&state).map_err(|e| e.to_string())
pub async fn toggle_mcp_app(
state: State<'_, AppState>,
server_id: String,
app: String,
enabled: bool,
) -> Result<(), String> {
let app_ty = AppType::from_str(&app).map_err(|e| e.to_string())?;
McpService::toggle_app(&state, &server_id, app_ty, enabled).map_err(|e| e.to_string())
}
+11 -4
View File
@@ -10,12 +10,12 @@ pub async fn open_external(app: AppHandle, url: String) -> Result<bool, String>
let url = if url.starts_with("http://") || url.starts_with("https://") {
url
} else {
format!("https://{}", url)
format!("https://{url}")
};
app.opener()
.open_url(&url, None::<String>)
.map_err(|e| format!("打开链接失败: {}", e))?;
.map_err(|e| format!("打开链接失败: {e}"))?;
Ok(true)
}
@@ -29,7 +29,7 @@ pub async fn check_for_updates(handle: AppHandle) -> Result<bool, String> {
"https://github.com/farion1231/cc-switch/releases/latest",
None::<String>,
)
.map_err(|e| format!("打开更新页面失败: {}", e))?;
.map_err(|e| format!("打开更新页面失败: {e}"))?;
Ok(true)
}
@@ -37,7 +37,7 @@ pub async fn check_for_updates(handle: AppHandle) -> Result<bool, String> {
/// 判断是否为便携版(绿色版)运行
#[tauri::command]
pub async fn is_portable_mode() -> Result<bool, String> {
let exe_path = std::env::current_exe().map_err(|e| format!("获取可执行路径失败: {}", e))?;
let exe_path = std::env::current_exe().map_err(|e| format!("获取可执行路径失败: {e}"))?;
if let Some(dir) = exe_path.parent() {
Ok(dir.join("portable.ini").is_file())
} else {
@@ -51,3 +51,10 @@ pub async fn is_portable_mode() -> Result<bool, String> {
pub async fn get_init_error() -> Result<Option<InitErrorPayload>, String> {
Ok(crate::init_status::get_init_error())
}
/// 获取 JSON→SQLite 迁移结果(若有)。
/// 只返回一次 true,之后返回 false,用于前端显示一次性 Toast 通知。
#[tauri::command]
pub async fn get_migration_result() -> Result<bool, String> {
Ok(crate::init_status::take_migration_success())
}
+14
View File
@@ -1,17 +1,31 @@
#![allow(non_snake_case)]
mod config;
mod deeplink;
mod env;
mod import_export;
mod mcp;
mod misc;
mod model_test;
mod plugin;
mod prompt;
mod provider;
mod proxy;
mod settings;
pub mod skill;
mod usage;
pub use config::*;
pub use deeplink::*;
pub use env::*;
pub use import_export::*;
pub use mcp::*;
pub use misc::*;
pub use model_test::*;
pub use plugin::*;
pub use prompt::*;
pub use provider::*;
pub use proxy::*;
pub use settings::*;
pub use skill::*;
pub use usage::*;
+128
View File
@@ -0,0 +1,128 @@
//! 模型测试相关命令
use crate::app_config::AppType;
use crate::error::AppError;
use crate::services::model_test::{
ModelTestConfig, ModelTestLog, ModelTestResult, ModelTestService,
};
use crate::store::AppState;
use tauri::State;
/// 测试单个供应商的模型可用性
#[tauri::command]
pub async fn test_provider_model(
state: State<'_, AppState>,
app_type: AppType,
provider_id: String,
) -> Result<ModelTestResult, AppError> {
// 获取测试配置
let config = state.db.get_model_test_config()?;
// 获取供应商
let providers = state.db.get_all_providers(app_type.as_str())?;
let provider = providers
.get(&provider_id)
.ok_or_else(|| AppError::Message(format!("供应商 {provider_id} 不存在")))?;
// 执行测试
let result = ModelTestService::test_provider(&app_type, provider, &config).await?;
// 记录日志
let _ = state.db.save_model_test_log(
&provider_id,
&provider.name,
app_type.as_str(),
&result.model_used,
&config.test_prompt,
&result,
);
Ok(result)
}
/// 批量测试所有供应商
#[tauri::command]
pub async fn test_all_providers_model(
state: State<'_, AppState>,
app_type: AppType,
proxy_targets_only: bool,
) -> Result<Vec<(String, ModelTestResult)>, AppError> {
let config = state.db.get_model_test_config()?;
let providers = state.db.get_all_providers(app_type.as_str())?;
let mut results = Vec::new();
for (id, provider) in providers {
// 如果只测试代理目标,跳过非代理目标
if proxy_targets_only && !provider.is_proxy_target.unwrap_or(false) {
continue;
}
match ModelTestService::test_provider(&app_type, &provider, &config).await {
Ok(result) => {
// 记录日志
let _ = state.db.save_model_test_log(
&id,
&provider.name,
app_type.as_str(),
&result.model_used,
&config.test_prompt,
&result,
);
results.push((id, result));
}
Err(e) => {
let error_result = ModelTestResult {
success: false,
message: e.to_string(),
response_time_ms: None,
http_status: None,
model_used: String::new(),
tested_at: chrono::Utc::now().timestamp(),
};
results.push((id, error_result));
}
}
}
Ok(results)
}
/// 获取模型测试配置
#[tauri::command]
pub fn get_model_test_config(state: State<'_, AppState>) -> Result<ModelTestConfig, AppError> {
state.db.get_model_test_config()
}
/// 保存模型测试配置
#[tauri::command]
pub fn save_model_test_config(
state: State<'_, AppState>,
config: ModelTestConfig,
) -> Result<(), AppError> {
state.db.save_model_test_config(&config)
}
/// 获取模型测试日志
#[tauri::command]
pub fn get_model_test_logs(
state: State<'_, AppState>,
app_type: Option<String>,
provider_id: Option<String>,
limit: Option<u32>,
) -> Result<Vec<ModelTestLog>, AppError> {
state.db.get_model_test_logs(
app_type.as_deref(),
provider_id.as_deref(),
limit.unwrap_or(50),
)
}
/// 清理旧的测试日志
#[tauri::command]
pub fn cleanup_model_test_logs(
state: State<'_, AppState>,
keep_count: Option<u32>,
) -> Result<u64, AppError> {
state.db.cleanup_model_test_logs(keep_count.unwrap_or(100))
}
+64
View File
@@ -0,0 +1,64 @@
use indexmap::IndexMap;
use std::str::FromStr;
use tauri::State;
use crate::app_config::AppType;
use crate::prompt::Prompt;
use crate::services::PromptService;
use crate::store::AppState;
#[tauri::command]
pub async fn get_prompts(
app: String,
state: State<'_, AppState>,
) -> Result<IndexMap<String, Prompt>, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::get_prompts(&state, app_type).map_err(|e| e.to_string())
}
#[tauri::command]
pub async fn upsert_prompt(
app: String,
id: String,
prompt: Prompt,
state: State<'_, AppState>,
) -> Result<(), String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::upsert_prompt(&state, app_type, &id, prompt).map_err(|e| e.to_string())
}
#[tauri::command]
pub async fn delete_prompt(
app: String,
id: String,
state: State<'_, AppState>,
) -> Result<(), String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::delete_prompt(&state, app_type, &id).map_err(|e| e.to_string())
}
#[tauri::command]
pub async fn enable_prompt(
app: String,
id: String,
state: State<'_, AppState>,
) -> Result<(), String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::enable_prompt(&state, app_type, &id).map_err(|e| e.to_string())
}
#[tauri::command]
pub async fn import_prompt_from_file(
app: String,
state: State<'_, AppState>,
) -> Result<String, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::import_from_file(&state, app_type).map_err(|e| e.to_string())
}
#[tauri::command]
pub async fn get_current_prompt_file_content(app: String) -> Result<Option<String>, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
PromptService::get_current_file_content(app_type).map_err(|e| e.to_string())
}
+23 -7
View File
@@ -1,4 +1,4 @@
use std::collections::HashMap;
use indexmap::IndexMap;
use tauri::State;
use crate::app_config::AppType;
@@ -13,7 +13,7 @@ use std::str::FromStr;
pub fn get_providers(
state: State<'_, AppState>,
app: String,
) -> Result<HashMap<String, Provider>, String> {
) -> Result<IndexMap<String, Provider>, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
ProviderService::list(state.inner(), app_type).map_err(|e| e.to_string())
}
@@ -86,7 +86,20 @@ pub fn switch_provider(
.map_err(|e| e.to_string())
}
fn import_default_config_internal(state: &AppState, app_type: AppType) -> Result<(), AppError> {
/// 设置代理目标供应商
#[tauri::command]
pub fn set_proxy_target_provider(
state: State<'_, AppState>,
app: String,
id: String,
) -> Result<bool, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
ProviderService::set_proxy_target(state.inner(), app_type, &id)
.map(|_| true)
.map_err(|e| e.to_string())
}
fn import_default_config_internal(state: &AppState, app_type: AppType) -> Result<bool, AppError> {
ProviderService::import_default_config(state, app_type)
}
@@ -94,7 +107,7 @@ fn import_default_config_internal(state: &AppState, app_type: AppType) -> Result
pub fn import_default_config_test_hook(
state: &AppState,
app_type: AppType,
) -> Result<(), AppError> {
) -> Result<bool, AppError> {
import_default_config_internal(state, app_type)
}
@@ -102,9 +115,7 @@ pub fn import_default_config_test_hook(
#[tauri::command]
pub fn import_default_config(state: State<'_, AppState>, app: String) -> Result<bool, String> {
let app_type = AppType::from_str(&app).map_err(|e| e.to_string())?;
import_default_config_internal(&state, app_type)
.map(|_| true)
.map_err(Into::into)
import_default_config_internal(&state, app_type).map_err(Into::into)
}
/// 查询供应商用量
@@ -123,6 +134,7 @@ pub async fn queryProviderUsage(
/// 测试用量脚本(使用当前编辑器中的脚本,不保存)
#[allow(non_snake_case)]
#[allow(clippy::too_many_arguments)]
#[tauri::command]
pub async fn testUsageScript(
state: State<'_, AppState>,
@@ -130,6 +142,8 @@ pub async fn testUsageScript(
app: String,
#[allow(non_snake_case)] scriptCode: String,
timeout: Option<u64>,
#[allow(non_snake_case)] apiKey: Option<String>,
#[allow(non_snake_case)] baseUrl: Option<String>,
#[allow(non_snake_case)] accessToken: Option<String>,
#[allow(non_snake_case)] userId: Option<String>,
) -> Result<crate::provider::UsageResult, String> {
@@ -140,6 +154,8 @@ pub async fn testUsageScript(
&providerId,
&scriptCode,
timeout.unwrap_or(10),
apiKey.as_deref(),
baseUrl.as_deref(),
accessToken.as_deref(),
userId.as_deref(),
)
+47
View File
@@ -0,0 +1,47 @@
//! 代理服务相关的 Tauri 命令
//!
//! 提供前端调用的 API 接口
use crate::proxy::types::*;
use crate::store::AppState;
/// 启动代理服务器
#[tauri::command]
pub async fn start_proxy_server(
state: tauri::State<'_, AppState>,
) -> Result<ProxyServerInfo, String> {
state.proxy_service.start().await
}
/// 停止代理服务器
#[tauri::command]
pub async fn stop_proxy_server(state: tauri::State<'_, AppState>) -> Result<(), String> {
state.proxy_service.stop().await
}
/// 获取代理服务器状态
#[tauri::command]
pub async fn get_proxy_status(state: tauri::State<'_, AppState>) -> Result<ProxyStatus, String> {
state.proxy_service.get_status().await
}
/// 获取代理配置
#[tauri::command]
pub async fn get_proxy_config(state: tauri::State<'_, AppState>) -> Result<ProxyConfig, String> {
state.proxy_service.get_config().await
}
/// 更新代理配置
#[tauri::command]
pub async fn update_proxy_config(
state: tauri::State<'_, AppState>,
config: ProxyConfig,
) -> Result<(), String> {
state.proxy_service.update_config(&config).await
}
/// 检查代理服务器是否正在运行
#[tauri::command]
pub async fn is_proxy_running(state: tauri::State<'_, AppState>) -> Result<bool, String> {
Ok(state.proxy_service.is_running().await)
}
+23 -1
View File
@@ -18,7 +18,12 @@ pub async fn save_settings(settings: crate::settings::AppSettings) -> Result<boo
/// 重启应用程序(当 app_config_dir 变更后使用)
#[tauri::command]
pub async fn restart_app(app: AppHandle) -> Result<bool, String> {
app.restart();
// 在后台延迟重启,让函数有时间返回响应
tauri::async_runtime::spawn(async move {
tokio::time::sleep(tokio::time::Duration::from_millis(100)).await;
app.restart();
});
Ok(true)
}
/// 获取 app_config_dir 覆盖配置 (从 Store)
@@ -37,3 +42,20 @@ pub async fn set_app_config_dir_override(
crate::app_store::set_app_config_dir_to_store(&app, path.as_deref())?;
Ok(true)
}
/// 设置开机自启
#[tauri::command]
pub async fn set_auto_launch(enabled: bool) -> Result<bool, String> {
if enabled {
crate::auto_launch::enable_auto_launch().map_err(|e| format!("启用开机自启失败: {e}"))?;
} else {
crate::auto_launch::disable_auto_launch().map_err(|e| format!("禁用开机自启失败: {e}"))?;
}
Ok(true)
}
/// 获取开机自启状态
#[tauri::command]
pub async fn get_auto_launch_status() -> Result<bool, String> {
crate::auto_launch::is_auto_launch_enabled().map_err(|e| format!("获取开机自启状态失败: {e}"))
}
+235
View File
@@ -0,0 +1,235 @@
use crate::app_config::AppType;
use crate::error::format_skill_error;
use crate::services::skill::SkillState;
use crate::services::{Skill, SkillRepo, SkillService};
use crate::store::AppState;
use chrono::Utc;
use std::sync::Arc;
use tauri::State;
pub struct SkillServiceState(pub Arc<SkillService>);
/// 解析 app 参数为 AppType
fn parse_app_type(app: &str) -> Result<AppType, String> {
match app.to_lowercase().as_str() {
"claude" => Ok(AppType::Claude),
"codex" => Ok(AppType::Codex),
"gemini" => Ok(AppType::Gemini),
_ => Err(format!("不支持的 app 类型: {app}")),
}
}
/// 根据 app_type 生成带前缀的 skill key
fn get_skill_key(app_type: &AppType, directory: &str) -> String {
let prefix = match app_type {
AppType::Claude => "claude",
AppType::Codex => "codex",
AppType::Gemini => "gemini",
};
format!("{prefix}:{directory}")
}
#[tauri::command]
pub async fn get_skills(
service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<Vec<Skill>, String> {
get_skills_for_app("claude".to_string(), service, app_state).await
}
#[tauri::command]
pub async fn get_skills_for_app(
app: String,
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<Vec<Skill>, String> {
let app_type = parse_app_type(&app)?;
let service = SkillService::new_for_app(app_type.clone()).map_err(|e| e.to_string())?;
let repos = app_state.db.get_skill_repos().map_err(|e| e.to_string())?;
let skills = service
.list_skills(repos)
.await
.map_err(|e| e.to_string())?;
// 自动同步本地已安装的 skills 到数据库
// 这样用户在首次运行时,已有的 skills 会被自动记录
let existing_states = app_state.db.get_skills().unwrap_or_default();
for skill in &skills {
if skill.installed {
let key = get_skill_key(&app_type, &skill.directory);
if !existing_states.contains_key(&key) {
// 本地有该 skill,但数据库中没有记录,自动添加
if let Err(e) = app_state.db.update_skill_state(
&key,
&SkillState {
installed: true,
installed_at: Utc::now(),
},
) {
log::warn!("同步本地 skill {key} 状态到数据库失败: {e}");
}
}
}
}
Ok(skills)
}
#[tauri::command]
pub async fn install_skill(
directory: String,
service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
install_skill_for_app("claude".to_string(), directory, service, app_state).await
}
#[tauri::command]
pub async fn install_skill_for_app(
app: String,
directory: String,
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
let app_type = parse_app_type(&app)?;
let service = SkillService::new_for_app(app_type.clone()).map_err(|e| e.to_string())?;
// 先在不持有写锁的情况下收集仓库与技能信息
let repos = app_state.db.get_skill_repos().map_err(|e| e.to_string())?;
let skills = service
.list_skills(repos)
.await
.map_err(|e| e.to_string())?;
let skill = skills
.iter()
.find(|s| s.directory.eq_ignore_ascii_case(&directory))
.ok_or_else(|| {
format_skill_error(
"SKILL_NOT_FOUND",
&[("directory", &directory)],
Some("checkRepoUrl"),
)
})?;
if !skill.installed {
let repo = SkillRepo {
owner: skill.repo_owner.clone().ok_or_else(|| {
format_skill_error(
"MISSING_REPO_INFO",
&[("directory", &directory), ("field", "owner")],
None,
)
})?,
name: skill.repo_name.clone().ok_or_else(|| {
format_skill_error(
"MISSING_REPO_INFO",
&[("directory", &directory), ("field", "name")],
None,
)
})?,
branch: skill
.repo_branch
.clone()
.unwrap_or_else(|| "main".to_string()),
enabled: true,
};
service
.install_skill(directory.clone(), repo)
.await
.map_err(|e| e.to_string())?;
}
let key = get_skill_key(&app_type, &directory);
app_state
.db
.update_skill_state(
&key,
&SkillState {
installed: true,
installed_at: Utc::now(),
},
)
.map_err(|e| e.to_string())?;
Ok(true)
}
#[tauri::command]
pub fn uninstall_skill(
directory: String,
service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
uninstall_skill_for_app("claude".to_string(), directory, service, app_state)
}
#[tauri::command]
pub fn uninstall_skill_for_app(
app: String,
directory: String,
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
let app_type = parse_app_type(&app)?;
let service = SkillService::new_for_app(app_type.clone()).map_err(|e| e.to_string())?;
service
.uninstall_skill(directory.clone())
.map_err(|e| e.to_string())?;
// Remove from database by setting installed = false
let key = get_skill_key(&app_type, &directory);
app_state
.db
.update_skill_state(
&key,
&SkillState {
installed: false,
installed_at: Utc::now(),
},
)
.map_err(|e| e.to_string())?;
Ok(true)
}
#[tauri::command]
pub fn get_skill_repos(
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<Vec<SkillRepo>, String> {
app_state.db.get_skill_repos().map_err(|e| e.to_string())
}
#[tauri::command]
pub fn add_skill_repo(
repo: SkillRepo,
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
app_state
.db
.save_skill_repo(&repo)
.map_err(|e| e.to_string())?;
Ok(true)
}
#[tauri::command]
pub fn remove_skill_repo(
owner: String,
name: String,
_service: State<'_, SkillServiceState>,
app_state: State<'_, AppState>,
) -> Result<bool, String> {
app_state
.db
.delete_skill_repo(&owner, &name)
.map_err(|e| e.to_string())?;
Ok(true)
}
+178
View File
@@ -0,0 +1,178 @@
//! 使用统计相关命令
use crate::error::AppError;
use crate::services::usage_stats::*;
use crate::store::AppState;
use tauri::State;
/// 获取使用量汇总
#[tauri::command]
pub fn get_usage_summary(
state: State<'_, AppState>,
start_date: Option<i64>,
end_date: Option<i64>,
) -> Result<UsageSummary, AppError> {
state.db.get_usage_summary(start_date, end_date)
}
/// 获取每日趋势
#[tauri::command]
pub fn get_usage_trends(
state: State<'_, AppState>,
days: u32,
) -> Result<Vec<DailyStats>, AppError> {
state.db.get_daily_trends(days)
}
/// 获取 Provider 统计
#[tauri::command]
pub fn get_provider_stats(state: State<'_, AppState>) -> Result<Vec<ProviderStats>, AppError> {
state.db.get_provider_stats()
}
/// 获取模型统计
#[tauri::command]
pub fn get_model_stats(state: State<'_, AppState>) -> Result<Vec<ModelStats>, AppError> {
state.db.get_model_stats()
}
/// 获取请求日志列表
#[tauri::command]
pub fn get_request_logs(
state: State<'_, AppState>,
filters: LogFilters,
page: u32,
page_size: u32,
) -> Result<PaginatedLogs, AppError> {
state.db.get_request_logs(&filters, page, page_size)
}
/// 获取单个请求详情
#[tauri::command]
pub fn get_request_detail(
state: State<'_, AppState>,
request_id: String,
) -> Result<Option<RequestLogDetail>, AppError> {
state.db.get_request_detail(&request_id)
}
/// 获取模型定价列表
#[tauri::command]
pub fn get_model_pricing(state: State<'_, AppState>) -> Result<Vec<ModelPricingInfo>, AppError> {
log::info!("获取模型定价列表");
state.db.ensure_model_pricing_seeded()?;
let db = state.db.clone();
let conn = crate::database::lock_conn!(db.conn);
// 检查表是否存在
let table_exists: bool = conn
.query_row(
"SELECT COUNT(*) FROM sqlite_master WHERE type='table' AND name='model_pricing'",
[],
|row| row.get::<_, i64>(0).map(|count| count > 0),
)
.unwrap_or(false);
if !table_exists {
log::error!("model_pricing 表不存在,可能需要重启应用以触发数据库迁移");
return Ok(Vec::new());
}
let mut stmt = conn.prepare(
"SELECT model_id, display_name, input_cost_per_million, output_cost_per_million,
cache_read_cost_per_million, cache_creation_cost_per_million
FROM model_pricing
ORDER BY display_name",
)?;
let rows = stmt.query_map([], |row| {
Ok(ModelPricingInfo {
model_id: row.get(0)?,
display_name: row.get(1)?,
input_cost_per_million: row.get(2)?,
output_cost_per_million: row.get(3)?,
cache_read_cost_per_million: row.get(4)?,
cache_creation_cost_per_million: row.get(5)?,
})
})?;
let mut pricing = Vec::new();
for row in rows {
pricing.push(row?);
}
log::info!("成功获取 {} 条模型定价数据", pricing.len());
Ok(pricing)
}
/// 更新模型定价
#[tauri::command]
pub fn update_model_pricing(
state: State<'_, AppState>,
model_id: String,
display_name: String,
input_cost: String,
output_cost: String,
cache_read_cost: String,
cache_creation_cost: String,
) -> Result<(), AppError> {
let db = state.db.clone();
let conn = crate::database::lock_conn!(db.conn);
conn.execute(
"INSERT OR REPLACE INTO model_pricing (
model_id, display_name, input_cost_per_million, output_cost_per_million,
cache_read_cost_per_million, cache_creation_cost_per_million
) VALUES (?1, ?2, ?3, ?4, ?5, ?6)",
rusqlite::params![
model_id,
display_name,
input_cost,
output_cost,
cache_read_cost,
cache_creation_cost
],
)
.map_err(|e| AppError::Database(format!("更新模型定价失败: {e}")))?;
Ok(())
}
/// 检查 Provider 使用限额
#[tauri::command]
pub fn check_provider_limits(
state: State<'_, AppState>,
provider_id: String,
app_type: String,
) -> Result<crate::services::usage_stats::ProviderLimitStatus, AppError> {
state.db.check_provider_limits(&provider_id, &app_type)
}
/// 删除模型定价
#[tauri::command]
pub fn delete_model_pricing(state: State<'_, AppState>, model_id: String) -> Result<(), AppError> {
let db = state.db.clone();
let conn = crate::database::lock_conn!(db.conn);
conn.execute(
"DELETE FROM model_pricing WHERE model_id = ?1",
rusqlite::params![model_id],
)
.map_err(|e| AppError::Database(format!("删除模型定价失败: {e}")))?;
log::info!("已删除模型定价: {model_id}");
Ok(())
}
/// 模型定价信息
#[derive(Debug, Clone, serde::Serialize, serde::Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct ModelPricingInfo {
pub model_id: String,
pub display_name: String,
pub input_cost_per_million: String,
pub output_cost_per_million: String,
pub cache_read_cost_per_million: String,
pub cache_creation_cost_per_million: String,
}
+5 -3
View File
@@ -33,7 +33,7 @@ fn derive_mcp_path_from_override(dir: &Path) -> Option<PathBuf> {
return None;
}
let parent = dir.parent().unwrap_or_else(|| Path::new(""));
Some(parent.join(format!("{}.json", file_name)))
Some(parent.join(format!("{file_name}.json")))
}
/// 获取 Claude MCP 配置文件路径,若设置了目录覆盖则与覆盖目录同级
@@ -79,6 +79,7 @@ pub fn get_app_config_path() -> PathBuf {
}
/// 清理供应商名称,确保文件名安全
#[allow(dead_code)]
pub fn sanitize_provider_name(name: &str) -> String {
name.chars()
.map(|c| match c {
@@ -90,12 +91,13 @@ pub fn sanitize_provider_name(name: &str) -> String {
}
/// 获取供应商配置文件路径
#[allow(dead_code)]
pub fn get_provider_config_path(provider_id: &str, provider_name: Option<&str>) -> PathBuf {
let base_name = provider_name
.map(sanitize_provider_name)
.unwrap_or_else(|| sanitize_provider_name(provider_id));
get_claude_config_dir().join(format!("settings-{}.json", base_name))
get_claude_config_dir().join(format!("settings-{base_name}.json"))
}
/// 读取 JSON 配置文件
@@ -149,7 +151,7 @@ pub fn atomic_write(path: &Path, data: &[u8]) -> Result<(), AppError> {
.duration_since(std::time::UNIX_EPOCH)
.unwrap_or_default()
.as_nanos();
tmp.push(format!("{}.tmp.{}", file_name, ts));
tmp.push(format!("{file_name}.tmp.{ts}"));
{
let mut f = fs::File::create(&tmp).map_err(|e| AppError::io(&tmp, e))?;
+324
View File
@@ -0,0 +1,324 @@
//! 数据库备份和恢复
//!
//! 提供 SQL 导出/导入和二进制快照备份功能。
use super::{lock_conn, Database, DB_BACKUP_RETAIN};
use crate::config::get_app_config_dir;
use crate::error::AppError;
use chrono::Utc;
use rusqlite::backup::Backup;
use rusqlite::types::ValueRef;
use rusqlite::Connection;
use std::fs;
use std::path::{Path, PathBuf};
use tempfile::NamedTempFile;
impl Database {
/// 导出为 SQLite 兼容的 SQL 文本
pub fn export_sql(&self, target_path: &Path) -> Result<(), AppError> {
let snapshot = self.snapshot_to_memory()?;
let dump = Self::dump_sql(&snapshot)?;
if let Some(parent) = target_path.parent() {
fs::create_dir_all(parent).map_err(|e| AppError::io(parent, e))?;
}
crate::config::atomic_write(target_path, dump.as_bytes())
}
/// 从 SQL 文件导入,返回生成的备份 ID(若无备份则为空字符串)
pub fn import_sql(&self, source_path: &Path) -> Result<String, AppError> {
if !source_path.exists() {
return Err(AppError::InvalidInput(format!(
"SQL 文件不存在: {}",
source_path.display()
)));
}
let sql_raw = fs::read_to_string(source_path).map_err(|e| AppError::io(source_path, e))?;
let sql_content = Self::sanitize_import_sql(&sql_raw);
// 导入前备份现有数据库
let backup_path = self.backup_database_file()?;
// 在临时数据库执行导入,确保失败不会污染主库
let temp_file = NamedTempFile::new().map_err(|e| AppError::IoContext {
context: "创建临时数据库文件失败".to_string(),
source: e,
})?;
let temp_path = temp_file.path().to_path_buf();
let temp_conn =
Connection::open(&temp_path).map_err(|e| AppError::Database(e.to_string()))?;
temp_conn
.execute_batch(&sql_content)
.map_err(|e| AppError::Database(format!("执行 SQL 导入失败: {e}")))?;
// 补齐缺失表/索引并进行基础校验
Self::create_tables_on_conn(&temp_conn)?;
Self::apply_schema_migrations_on_conn(&temp_conn)?;
Self::validate_basic_state(&temp_conn)?;
// 使用 Backup 将临时库原子写回主库
{
let mut main_conn = lock_conn!(self.conn);
let backup = Backup::new(&temp_conn, &mut main_conn)
.map_err(|e| AppError::Database(e.to_string()))?;
backup
.step(-1)
.map_err(|e| AppError::Database(e.to_string()))?;
}
let backup_id = backup_path
.and_then(|p| p.file_stem().map(|s| s.to_string_lossy().to_string()))
.unwrap_or_default();
Ok(backup_id)
}
/// 创建内存快照以避免长时间持有数据库锁
pub(crate) fn snapshot_to_memory(&self) -> Result<Connection, AppError> {
let conn = lock_conn!(self.conn);
let mut snapshot =
Connection::open_in_memory().map_err(|e| AppError::Database(e.to_string()))?;
{
let backup =
Backup::new(&conn, &mut snapshot).map_err(|e| AppError::Database(e.to_string()))?;
backup
.step(-1)
.map_err(|e| AppError::Database(e.to_string()))?;
}
Ok(snapshot)
}
/// 移除 SQLite 保留对象相关语句(如 sqlite_sequence),避免导入报错
fn sanitize_import_sql(sql: &str) -> String {
let mut cleaned = String::new();
let lower_keyword = "sqlite_sequence";
for stmt in sql.split(';') {
let trimmed = stmt.trim();
if trimmed.is_empty() {
continue;
}
if trimmed.to_ascii_lowercase().contains(lower_keyword) {
continue;
}
cleaned.push_str(trimmed);
cleaned.push_str(";\n");
}
cleaned
}
/// 生成一致性快照备份,返回备份文件路径(不存在主库时返回 None)
fn backup_database_file(&self) -> Result<Option<PathBuf>, AppError> {
let db_path = get_app_config_dir().join("cc-switch.db");
if !db_path.exists() {
return Ok(None);
}
let backup_dir = db_path
.parent()
.ok_or_else(|| AppError::Config("无效的数据库路径".to_string()))?
.join("backups");
fs::create_dir_all(&backup_dir).map_err(|e| AppError::io(&backup_dir, e))?;
let backup_id = format!("db_backup_{}", Utc::now().format("%Y%m%d_%H%M%S"));
let backup_path = backup_dir.join(format!("{backup_id}.db"));
{
let conn = lock_conn!(self.conn);
let mut dest_conn =
Connection::open(&backup_path).map_err(|e| AppError::Database(e.to_string()))?;
let backup = Backup::new(&conn, &mut dest_conn)
.map_err(|e| AppError::Database(e.to_string()))?;
backup
.step(-1)
.map_err(|e| AppError::Database(e.to_string()))?;
}
Self::cleanup_db_backups(&backup_dir)?;
Ok(Some(backup_path))
}
/// 清理旧的数据库备份,保留最新的 N 个
fn cleanup_db_backups(dir: &Path) -> Result<(), AppError> {
let entries = match fs::read_dir(dir) {
Ok(iter) => iter
.filter_map(|entry| entry.ok())
.filter(|entry| {
entry
.path()
.extension()
.map(|ext| ext == "db")
.unwrap_or(false)
})
.collect::<Vec<_>>(),
Err(_) => return Ok(()),
};
if entries.len() <= DB_BACKUP_RETAIN {
return Ok(());
}
let remove_count = entries.len().saturating_sub(DB_BACKUP_RETAIN);
let mut sorted = entries;
sorted.sort_by_key(|entry| entry.metadata().and_then(|m| m.modified()).ok());
for entry in sorted.into_iter().take(remove_count) {
if let Err(err) = fs::remove_file(entry.path()) {
log::warn!("删除旧数据库备份失败 {}: {}", entry.path().display(), err);
}
}
Ok(())
}
/// 基础状态校验
fn validate_basic_state(conn: &Connection) -> Result<(), AppError> {
let provider_count: i64 = conn
.query_row("SELECT COUNT(*) FROM providers", [], |row| row.get(0))
.map_err(|e| AppError::Database(e.to_string()))?;
let mcp_count: i64 = conn
.query_row("SELECT COUNT(*) FROM mcp_servers", [], |row| row.get(0))
.map_err(|e| AppError::Database(e.to_string()))?;
if provider_count == 0 && mcp_count == 0 {
return Err(AppError::Config(
"导入的 SQL 未包含有效的供应商或 MCP 数据".to_string(),
));
}
Ok(())
}
/// 导出数据库为 SQL 文本
fn dump_sql(conn: &Connection) -> Result<String, AppError> {
let mut output = String::new();
let timestamp = Utc::now().format("%Y-%m-%d %H:%M:%S").to_string();
let user_version: i64 = conn
.query_row("PRAGMA user_version;", [], |row| row.get(0))
.unwrap_or(0);
output.push_str(&format!(
"-- CC Switch SQLite 导出\n-- 生成时间: {timestamp}\n-- user_version: {user_version}\n"
));
output.push_str("PRAGMA foreign_keys=OFF;\n");
output.push_str(&format!("PRAGMA user_version={user_version};\n"));
output.push_str("BEGIN TRANSACTION;\n");
// 导出 schema
let mut stmt = conn
.prepare(
"SELECT type, name, tbl_name, sql
FROM sqlite_master
WHERE sql NOT NULL AND type IN ('table','index','trigger','view')
ORDER BY type='table' DESC, name",
)
.map_err(|e| AppError::Database(e.to_string()))?;
let mut tables = Vec::new();
let mut rows = stmt
.query([])
.map_err(|e| AppError::Database(e.to_string()))?;
while let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
let obj_type: String = row.get(0).map_err(|e| AppError::Database(e.to_string()))?;
let name: String = row.get(1).map_err(|e| AppError::Database(e.to_string()))?;
let sql: String = row.get(3).map_err(|e| AppError::Database(e.to_string()))?;
// 跳过 SQLite 内部对象(如 sqlite_sequence
if name.starts_with("sqlite_") {
continue;
}
output.push_str(&sql);
output.push_str(";\n");
if obj_type == "table" && !name.starts_with("sqlite_") {
tables.push(name);
}
}
// 导出数据
for table in tables {
let columns = Self::get_table_columns(conn, &table)?;
if columns.is_empty() {
continue;
}
let mut stmt = conn
.prepare(&format!("SELECT * FROM \"{table}\""))
.map_err(|e| AppError::Database(e.to_string()))?;
let mut rows = stmt
.query([])
.map_err(|e| AppError::Database(e.to_string()))?;
while let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
let mut values = Vec::with_capacity(columns.len());
for idx in 0..columns.len() {
let value = row
.get_ref(idx)
.map_err(|e| AppError::Database(e.to_string()))?;
values.push(Self::format_sql_value(value)?);
}
let cols = columns
.iter()
.map(|c| format!("\"{c}\""))
.collect::<Vec<_>>()
.join(", ");
output.push_str(&format!(
"INSERT INTO \"{table}\" ({cols}) VALUES ({});\n",
values.join(", ")
));
}
}
output.push_str("COMMIT;\nPRAGMA foreign_keys=ON;\n");
Ok(output)
}
/// 获取表的列名列表
fn get_table_columns(conn: &Connection, table: &str) -> Result<Vec<String>, AppError> {
let mut stmt = conn
.prepare(&format!("PRAGMA table_info(\"{table}\")"))
.map_err(|e| AppError::Database(e.to_string()))?;
let iter = stmt
.query_map([], |row| row.get::<_, String>(1))
.map_err(|e| AppError::Database(e.to_string()))?;
let mut columns = Vec::new();
for col in iter {
columns.push(col.map_err(|e| AppError::Database(e.to_string()))?);
}
Ok(columns)
}
/// 格式化 SQL 值
fn format_sql_value(value: ValueRef<'_>) -> Result<String, AppError> {
match value {
ValueRef::Null => Ok("NULL".to_string()),
ValueRef::Integer(i) => Ok(i.to_string()),
ValueRef::Real(f) => Ok(f.to_string()),
ValueRef::Text(t) => {
let text = std::str::from_utf8(t)
.map_err(|e| AppError::Database(format!("文本字段不是有效的 UTF-8: {e}")))?;
let escaped = text.replace('\'', "''");
Ok(format!("'{escaped}'"))
}
ValueRef::Blob(bytes) => {
let mut s = String::from("X'");
for b in bytes {
use std::fmt::Write;
let _ = write!(&mut s, "{b:02X}");
}
s.push('\'');
Ok(s)
}
}
}
}
+97
View File
@@ -0,0 +1,97 @@
//! MCP 服务器数据访问对象
//!
//! 提供 MCP 服务器的 CRUD 操作。
use crate::app_config::{McpApps, McpServer};
use crate::database::{lock_conn, Database};
use crate::error::AppError;
use indexmap::IndexMap;
use rusqlite::params;
impl Database {
/// 获取所有 MCP 服务器
pub fn get_all_mcp_servers(&self) -> Result<IndexMap<String, McpServer>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn.prepare(
"SELECT id, name, server_config, description, homepage, docs, tags, enabled_claude, enabled_codex, enabled_gemini
FROM mcp_servers
ORDER BY name ASC, id ASC"
).map_err(|e| AppError::Database(e.to_string()))?;
let server_iter = stmt
.query_map([], |row| {
let id: String = row.get(0)?;
let name: String = row.get(1)?;
let server_config_str: String = row.get(2)?;
let description: Option<String> = row.get(3)?;
let homepage: Option<String> = row.get(4)?;
let docs: Option<String> = row.get(5)?;
let tags_str: String = row.get(6)?;
let enabled_claude: bool = row.get(7)?;
let enabled_codex: bool = row.get(8)?;
let enabled_gemini: bool = row.get(9)?;
let server = serde_json::from_str(&server_config_str).unwrap_or_default();
let tags = serde_json::from_str(&tags_str).unwrap_or_default();
Ok((
id.clone(),
McpServer {
id,
name,
server,
apps: McpApps {
claude: enabled_claude,
codex: enabled_codex,
gemini: enabled_gemini,
},
description,
homepage,
docs,
tags,
},
))
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut servers = IndexMap::new();
for server_res in server_iter {
let (id, server) = server_res.map_err(|e| AppError::Database(e.to_string()))?;
servers.insert(id, server);
}
Ok(servers)
}
/// 保存 MCP 服务器
pub fn save_mcp_server(&self, server: &McpServer) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO mcp_servers (
id, name, server_config, description, homepage, docs, tags,
enabled_claude, enabled_codex, enabled_gemini
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9, ?10)",
params![
server.id,
server.name,
serde_json::to_string(&server.server).unwrap(),
server.description,
server.homepage,
server.docs,
serde_json::to_string(&server.tags).unwrap(),
server.apps.claude,
server.apps.codex,
server.apps.gemini,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 删除 MCP 服务器
pub fn delete_mcp_server(&self, id: &str) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute("DELETE FROM mcp_servers WHERE id = ?1", params![id])
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
}
+12
View File
@@ -0,0 +1,12 @@
//! Data Access Object layer
//!
//! Database access operations for each domain
pub mod mcp;
pub mod prompts;
pub mod providers;
pub mod proxy;
pub mod settings;
pub mod skills;
// 所有 DAO 方法都通过 Database impl 提供,无需单独导出
+88
View File
@@ -0,0 +1,88 @@
//! 提示词数据访问对象
//!
//! 提供提示词(Prompt)的 CRUD 操作。
use crate::database::{lock_conn, Database};
use crate::error::AppError;
use crate::prompt::Prompt;
use indexmap::IndexMap;
use rusqlite::params;
impl Database {
/// 获取指定应用类型的所有提示词
pub fn get_prompts(&self, app_type: &str) -> Result<IndexMap<String, Prompt>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare(
"SELECT id, name, content, description, enabled, created_at, updated_at
FROM prompts WHERE app_type = ?1
ORDER BY created_at ASC, id ASC",
)
.map_err(|e| AppError::Database(e.to_string()))?;
let prompt_iter = stmt
.query_map(params![app_type], |row| {
let id: String = row.get(0)?;
let name: String = row.get(1)?;
let content: String = row.get(2)?;
let description: Option<String> = row.get(3)?;
let enabled: bool = row.get(4)?;
let created_at: Option<i64> = row.get(5)?;
let updated_at: Option<i64> = row.get(6)?;
Ok((
id.clone(),
Prompt {
id,
name,
content,
description,
enabled,
created_at,
updated_at,
},
))
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut prompts = IndexMap::new();
for prompt_res in prompt_iter {
let (id, prompt) = prompt_res.map_err(|e| AppError::Database(e.to_string()))?;
prompts.insert(id, prompt);
}
Ok(prompts)
}
/// 保存提示词
pub fn save_prompt(&self, app_type: &str, prompt: &Prompt) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO prompts (
id, app_type, name, content, description, enabled, created_at, updated_at
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8)",
params![
prompt.id,
app_type,
prompt.name,
prompt.content,
prompt.description,
prompt.enabled,
prompt.created_at,
prompt.updated_at,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 删除提示词
pub fn delete_prompt(&self, app_type: &str, id: &str) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"DELETE FROM prompts WHERE id = ?1 AND app_type = ?2",
params![id, app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
}
+407
View File
@@ -0,0 +1,407 @@
//! 供应商数据访问对象
//!
//! 提供供应商(Provider)的 CRUD 操作。
use crate::database::{lock_conn, Database};
use crate::error::AppError;
use crate::provider::{Provider, ProviderMeta};
use indexmap::IndexMap;
use rusqlite::params;
use std::collections::HashMap;
impl Database {
/// 获取指定应用类型的所有供应商
pub fn get_all_providers(
&self,
app_type: &str,
) -> Result<IndexMap<String, Provider>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn.prepare(
"SELECT id, name, settings_config, website_url, category, created_at, sort_index, notes, icon, icon_color, meta, is_proxy_target
FROM providers WHERE app_type = ?1
ORDER BY COALESCE(sort_index, 999999), created_at ASC, id ASC"
).map_err(|e| AppError::Database(e.to_string()))?;
let provider_iter = stmt
.query_map(params![app_type], |row| {
let id: String = row.get(0)?;
let name: String = row.get(1)?;
let settings_config_str: String = row.get(2)?;
let website_url: Option<String> = row.get(3)?;
let category: Option<String> = row.get(4)?;
let created_at: Option<i64> = row.get(5)?;
let sort_index: Option<usize> = row.get(6)?;
let notes: Option<String> = row.get(7)?;
let icon: Option<String> = row.get(8)?;
let icon_color: Option<String> = row.get(9)?;
let meta_str: String = row.get(10)?;
let is_proxy_target: bool = row.get(11)?;
let settings_config =
serde_json::from_str(&settings_config_str).unwrap_or(serde_json::Value::Null);
let meta: ProviderMeta = serde_json::from_str(&meta_str).unwrap_or_default();
Ok((
id,
Provider {
id: "".to_string(), // Placeholder, set below
name,
settings_config,
website_url,
category,
created_at,
sort_index,
notes,
meta: Some(meta),
icon,
icon_color,
is_proxy_target: Some(is_proxy_target),
},
))
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut providers = IndexMap::new();
for provider_res in provider_iter {
let (id, mut provider) = provider_res.map_err(|e| AppError::Database(e.to_string()))?;
provider.id = id.clone();
// 加载 endpoints
let mut stmt_endpoints = conn.prepare(
"SELECT url, added_at FROM provider_endpoints WHERE provider_id = ?1 AND app_type = ?2 ORDER BY added_at ASC, url ASC"
).map_err(|e| AppError::Database(e.to_string()))?;
let endpoints_iter = stmt_endpoints
.query_map(params![id, app_type], |row| {
let url: String = row.get(0)?;
let added_at: Option<i64> = row.get(1)?;
Ok((
url,
crate::settings::CustomEndpoint {
url: "".to_string(),
added_at: added_at.unwrap_or(0),
last_used: None,
},
))
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut custom_endpoints = HashMap::new();
for ep_res in endpoints_iter {
let (url, mut ep) = ep_res.map_err(|e| AppError::Database(e.to_string()))?;
ep.url = url.clone();
custom_endpoints.insert(url, ep);
}
if let Some(meta) = &mut provider.meta {
meta.custom_endpoints = custom_endpoints;
}
providers.insert(id, provider);
}
Ok(providers)
}
/// 获取当前激活的供应商 ID
pub fn get_current_provider(&self, app_type: &str) -> Result<Option<String>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare("SELECT id FROM providers WHERE app_type = ?1 AND is_current = 1 LIMIT 1")
.map_err(|e| AppError::Database(e.to_string()))?;
let mut rows = stmt
.query(params![app_type])
.map_err(|e| AppError::Database(e.to_string()))?;
if let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
Ok(Some(
row.get(0).map_err(|e| AppError::Database(e.to_string()))?,
))
} else {
Ok(None)
}
}
/// 根据 ID 获取单个供应商
pub fn get_provider_by_id(
&self,
id: &str,
app_type: &str,
) -> Result<Option<Provider>, AppError> {
let conn = lock_conn!(self.conn);
let result = conn.query_row(
"SELECT name, settings_config, website_url, category, created_at, sort_index, notes, icon, icon_color, meta, is_proxy_target
FROM providers WHERE id = ?1 AND app_type = ?2",
params![id, app_type],
|row| {
let name: String = row.get(0)?;
let settings_config_str: String = row.get(1)?;
let website_url: Option<String> = row.get(2)?;
let category: Option<String> = row.get(3)?;
let created_at: Option<i64> = row.get(4)?;
let sort_index: Option<usize> = row.get(5)?;
let notes: Option<String> = row.get(6)?;
let icon: Option<String> = row.get(7)?;
let icon_color: Option<String> = row.get(8)?;
let meta_str: String = row.get(9)?;
let is_proxy_target: bool = row.get(10)?;
let settings_config = serde_json::from_str(&settings_config_str).unwrap_or(serde_json::Value::Null);
let meta: ProviderMeta = serde_json::from_str(&meta_str).unwrap_or_default();
Ok(Provider {
id: id.to_string(),
name,
settings_config,
website_url,
category,
created_at,
sort_index,
notes,
meta: Some(meta),
icon,
icon_color,
is_proxy_target: Some(is_proxy_target),
})
},
);
match result {
Ok(provider) => Ok(Some(provider)),
Err(rusqlite::Error::QueryReturnedNoRows) => Ok(None),
Err(e) => Err(AppError::Database(e.to_string())),
}
}
/// 获取代理目标供应商 ID
pub fn get_proxy_target_provider(&self, app_type: &str) -> Result<Option<String>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare("SELECT id FROM providers WHERE app_type = ?1 AND is_proxy_target = 1 LIMIT 1")
.map_err(|e| AppError::Database(e.to_string()))?;
let mut rows = stmt
.query(params![app_type])
.map_err(|e| AppError::Database(e.to_string()))?;
if let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
Ok(Some(
row.get(0).map_err(|e| AppError::Database(e.to_string()))?,
))
} else {
Ok(None)
}
}
/// 保存供应商(新增或更新)
///
/// 注意:更新模式下不同步 endpoints,因为编辑模式下端点通过单独的 API 管理
/// add_custom_endpoint / remove_custom_endpoint),避免覆盖用户的修改。
pub fn save_provider(&self, app_type: &str, provider: &Provider) -> Result<(), AppError> {
let mut conn = lock_conn!(self.conn);
let tx = conn
.transaction()
.map_err(|e| AppError::Database(e.to_string()))?;
// 处理 meta:取出 endpoints 以便单独处理
let mut meta_clone = provider.meta.clone().unwrap_or_default();
let endpoints = std::mem::take(&mut meta_clone.custom_endpoints);
// 检查是否存在(用于判断新增/更新,以及保留 is_current 和 is_proxy_target
let existing: Option<(bool, bool)> = tx
.query_row(
"SELECT is_current, is_proxy_target FROM providers WHERE id = ?1 AND app_type = ?2",
params![provider.id, app_type],
|row| Ok((row.get(0)?, row.get(1)?)),
)
.ok();
let is_update = existing.is_some();
let (is_current, is_proxy_target) = existing.unwrap_or((false, false));
if is_update {
// 更新模式:使用 UPDATE 避免触发 ON DELETE CASCADE
tx.execute(
"UPDATE providers SET
name = ?1,
settings_config = ?2,
website_url = ?3,
category = ?4,
created_at = ?5,
sort_index = ?6,
notes = ?7,
icon = ?8,
icon_color = ?9,
meta = ?10,
is_current = ?11,
is_proxy_target = ?12
WHERE id = ?13 AND app_type = ?14",
params![
provider.name,
serde_json::to_string(&provider.settings_config).unwrap(),
provider.website_url,
provider.category,
provider.created_at,
provider.sort_index,
provider.notes,
provider.icon,
provider.icon_color,
serde_json::to_string(&meta_clone).unwrap(),
is_current,
is_proxy_target,
provider.id,
app_type,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
} else {
// 新增模式:使用 INSERT
tx.execute(
"INSERT INTO providers (
id, app_type, name, settings_config, website_url, category,
created_at, sort_index, notes, icon, icon_color, meta, is_current, is_proxy_target
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9, ?10, ?11, ?12, ?13, ?14)",
params![
provider.id,
app_type,
provider.name,
serde_json::to_string(&provider.settings_config).unwrap(),
provider.website_url,
provider.category,
provider.created_at,
provider.sort_index,
provider.notes,
provider.icon,
provider.icon_color,
serde_json::to_string(&meta_clone).unwrap(),
is_current,
is_proxy_target,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 只有新增时才同步 endpoints
for (url, endpoint) in endpoints {
tx.execute(
"INSERT INTO provider_endpoints (provider_id, app_type, url, added_at)
VALUES (?1, ?2, ?3, ?4)",
params![provider.id, app_type, url, endpoint.added_at],
)
.map_err(|e| AppError::Database(e.to_string()))?;
}
}
tx.commit().map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 删除供应商
pub fn delete_provider(&self, app_type: &str, id: &str) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"DELETE FROM providers WHERE id = ?1 AND app_type = ?2",
params![id, app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 设置当前供应商
pub fn set_current_provider(&self, app_type: &str, id: &str) -> Result<(), AppError> {
let mut conn = lock_conn!(self.conn);
let tx = conn
.transaction()
.map_err(|e| AppError::Database(e.to_string()))?;
// 重置所有为 0
tx.execute(
"UPDATE providers SET is_current = 0 WHERE app_type = ?1",
params![app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 设置新的当前供应商
tx.execute(
"UPDATE providers SET is_current = 1 WHERE id = ?1 AND app_type = ?2",
params![id, app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
tx.commit().map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 设置代理目标供应商
pub fn set_proxy_target_provider(&self, app_type: &str, id: &str) -> Result<(), AppError> {
let mut conn = lock_conn!(self.conn);
let tx = conn
.transaction()
.map_err(|e| AppError::Database(e.to_string()))?;
// 重置所有为 0
tx.execute(
"UPDATE providers SET is_proxy_target = 0 WHERE app_type = ?1",
params![app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 设置新的代理目标供应商
tx.execute(
"UPDATE providers SET is_proxy_target = 1 WHERE id = ?1 AND app_type = ?2",
params![id, app_type],
)
.map_err(|e| AppError::Database(e.to_string()))?;
tx.commit().map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 获取所有活跃的代理目标
pub fn get_all_proxy_targets(&self) -> Result<Vec<(String, String, String)>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare("SELECT app_type, name, id FROM providers WHERE is_proxy_target = 1")
.map_err(|e| AppError::Database(e.to_string()))?;
let targets = stmt
.query_map([], |row| Ok((row.get(0)?, row.get(1)?, row.get(2)?)))
.map_err(|e| AppError::Database(e.to_string()))?
.collect::<Result<Vec<_>, _>>()
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(targets)
}
/// 添加自定义端点
pub fn add_custom_endpoint(
&self,
app_type: &str,
provider_id: &str,
url: &str,
) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
let added_at = chrono::Utc::now().timestamp_millis();
conn.execute(
"INSERT INTO provider_endpoints (provider_id, app_type, url, added_at) VALUES (?1, ?2, ?3, ?4)",
params![provider_id, app_type, url, added_at],
).map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 移除自定义端点
pub fn remove_custom_endpoint(
&self,
app_type: &str,
provider_id: &str,
url: &str,
) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"DELETE FROM provider_endpoints WHERE provider_id = ?1 AND app_type = ?2 AND url = ?3",
params![provider_id, app_type, url],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
}
+244
View File
@@ -0,0 +1,244 @@
//! 代理功能数据访问层
//!
//! 处理代理配置、Provider健康状态和使用统计的数据库操作
use crate::error::AppError;
use crate::proxy::types::*;
use super::super::{lock_conn, Database};
impl Database {
// ==================== Proxy Config ====================
/// 获取代理配置
pub async fn get_proxy_config(&self) -> Result<ProxyConfig, AppError> {
// 在一个作用域内获取锁并查询,确保锁在await之前释放
let result = {
let conn = lock_conn!(self.conn);
conn.query_row(
"SELECT enabled, listen_address, listen_port, max_retries,
request_timeout, enable_logging
FROM proxy_config WHERE id = 1",
[],
|row| {
Ok(ProxyConfig {
enabled: row.get::<_, i32>(0)? != 0,
listen_address: row.get(1)?,
listen_port: row.get::<_, i32>(2)? as u16,
max_retries: row.get::<_, i32>(3)? as u8,
request_timeout: row.get::<_, i32>(4)? as u64,
enable_logging: row.get::<_, i32>(5)? != 0,
})
},
)
}; // conn锁在这里释放
match result {
Ok(config) => Ok(config),
Err(rusqlite::Error::QueryReturnedNoRows) => {
// 如果不存在,插入默认配置
let default_config = ProxyConfig::default();
self.update_proxy_config(default_config.clone()).await?;
Ok(default_config)
}
Err(e) => Err(AppError::Database(e.to_string())),
}
}
/// 更新代理配置
pub async fn update_proxy_config(&self, config: ProxyConfig) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO proxy_config
(id, enabled, listen_address, listen_port, max_retries, request_timeout, enable_logging, target_app, created_at, updated_at)
VALUES (1, ?1, ?2, ?3, ?4, ?5, ?6, ?7,
COALESCE((SELECT created_at FROM proxy_config WHERE id = 1), datetime('now')),
datetime('now'))",
rusqlite::params![
if config.enabled { 1 } else { 0 },
config.listen_address,
config.listen_port as i32,
config.max_retries as i32,
config.request_timeout as i32,
if config.enable_logging { 1 } else { 0 },
"claude", // 兼容旧字段,写入默认值
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
// ==================== Provider Health ====================
/// 获取Provider健康状态
pub async fn get_provider_health(
&self,
provider_id: &str,
app_type: &str,
) -> Result<ProviderHealth, AppError> {
let conn = lock_conn!(self.conn);
conn.query_row(
"SELECT provider_id, app_type, is_healthy, consecutive_failures,
last_success_at, last_failure_at, last_error, updated_at
FROM provider_health
WHERE provider_id = ?1 AND app_type = ?2",
rusqlite::params![provider_id, app_type],
|row| {
Ok(ProviderHealth {
provider_id: row.get(0)?,
app_type: row.get(1)?,
is_healthy: row.get::<_, i64>(2)? != 0,
consecutive_failures: row.get::<_, i64>(3)? as u32,
last_success_at: row.get(4)?,
last_failure_at: row.get(5)?,
last_error: row.get(6)?,
updated_at: row.get(7)?,
})
},
)
.map_err(|e| AppError::Database(e.to_string()))
}
/// 更新Provider健康状态
pub async fn update_provider_health(
&self,
provider_id: &str,
app_type: &str,
success: bool,
error_msg: Option<String>,
) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
let now = chrono::Utc::now().to_rfc3339();
// 先查询当前状态
let current = conn.query_row(
"SELECT consecutive_failures FROM provider_health
WHERE provider_id = ?1 AND app_type = ?2",
rusqlite::params![provider_id, app_type],
|row| Ok(row.get::<_, i64>(0)? as u32),
);
let (is_healthy, consecutive_failures) = if success {
// 成功:重置失败计数
(1, 0)
} else {
// 失败:增加失败计数
let failures = current.unwrap_or(0) + 1;
let healthy = if failures >= 3 { 0 } else { 1 };
(healthy, failures)
};
let (last_success_at, last_failure_at) = if success {
(Some(now.clone()), None)
} else {
(None, Some(now.clone()))
};
// UPSERT
conn.execute(
"INSERT OR REPLACE INTO provider_health
(provider_id, app_type, is_healthy, consecutive_failures,
last_success_at, last_failure_at, last_error, updated_at)
VALUES (?1, ?2, ?3, ?4,
COALESCE(?5, (SELECT last_success_at FROM provider_health
WHERE provider_id = ?1 AND app_type = ?2)),
COALESCE(?6, (SELECT last_failure_at FROM provider_health
WHERE provider_id = ?1 AND app_type = ?2)),
?7, ?8)",
rusqlite::params![
provider_id,
app_type,
is_healthy,
consecutive_failures as i64,
last_success_at,
last_failure_at,
error_msg,
&now,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
// ==================== Proxy Usage (可选) ====================
/// 记录代理使用统计
#[allow(dead_code)]
pub async fn record_proxy_usage(&self, record: &ProxyUsageRecord) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT INTO proxy_usage
(provider_id, app_type, endpoint, request_tokens, response_tokens,
status_code, latency_ms, error, timestamp)
VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
rusqlite::params![
&record.provider_id,
&record.app_type,
&record.endpoint,
record.request_tokens,
record.response_tokens,
record.status_code as i64,
record.latency_ms as i64,
&record.error,
&record.timestamp,
],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 查询最近的使用统计
#[allow(dead_code)]
pub async fn get_recent_usage(
&self,
provider_id: &str,
app_type: &str,
limit: usize,
) -> Result<Vec<ProxyUsageRecord>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare(
"SELECT provider_id, app_type, endpoint, request_tokens, response_tokens,
status_code, latency_ms, error, timestamp
FROM proxy_usage
WHERE provider_id = ?1 AND app_type = ?2
ORDER BY timestamp DESC
LIMIT ?3",
)
.map_err(|e| AppError::Database(e.to_string()))?;
let rows = stmt
.query_map(
rusqlite::params![provider_id, app_type, limit as i64],
|row| {
Ok(ProxyUsageRecord {
provider_id: row.get(0)?,
app_type: row.get(1)?,
endpoint: row.get(2)?,
request_tokens: row.get(3)?,
response_tokens: row.get(4)?,
status_code: row.get::<_, i64>(5)? as u16,
latency_ms: row.get::<_, i64>(6)? as u64,
error: row.get(7)?,
timestamp: row.get(8)?,
})
},
)
.map_err(|e| AppError::Database(e.to_string()))?;
let mut records = Vec::new();
for row in rows {
records.push(row.map_err(|e| AppError::Database(e.to_string()))?);
}
Ok(records)
}
}
+65
View File
@@ -0,0 +1,65 @@
//! 通用设置数据访问对象
//!
//! 提供键值对形式的通用设置存储。
use crate::database::{lock_conn, Database};
use crate::error::AppError;
use rusqlite::params;
impl Database {
/// 获取设置值
pub fn get_setting(&self, key: &str) -> Result<Option<String>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare("SELECT value FROM settings WHERE key = ?1")
.map_err(|e| AppError::Database(e.to_string()))?;
let mut rows = stmt
.query(params![key])
.map_err(|e| AppError::Database(e.to_string()))?;
if let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
Ok(Some(
row.get(0).map_err(|e| AppError::Database(e.to_string()))?,
))
} else {
Ok(None)
}
}
/// 设置值
pub fn set_setting(&self, key: &str, value: &str) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO settings (key, value) VALUES (?1, ?2)",
params![key, value],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
// --- Config Snippets 辅助方法 ---
/// 获取通用配置片段
pub fn get_config_snippet(&self, app_type: &str) -> Result<Option<String>, AppError> {
self.get_setting(&format!("common_config_{app_type}"))
}
/// 设置通用配置片段
pub fn set_config_snippet(
&self,
app_type: &str,
snippet: Option<String>,
) -> Result<(), AppError> {
let key = format!("common_config_{app_type}");
if let Some(value) = snippet {
self.set_setting(&key, &value)
} else {
// 如果为 None 则删除
let conn = lock_conn!(self.conn);
conn.execute("DELETE FROM settings WHERE key = ?1", params![key])
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
}
}
+139
View File
@@ -0,0 +1,139 @@
//! Skills 数据访问对象
//!
//! 提供 Skills 和 Skill Repos 的 CRUD 操作。
use crate::database::{lock_conn, Database};
use crate::error::AppError;
use crate::services::skill::{SkillRepo, SkillState};
use indexmap::IndexMap;
use rusqlite::params;
impl Database {
/// 获取所有 Skills 状态
pub fn get_skills(&self) -> Result<IndexMap<String, SkillState>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare("SELECT directory, app_type, installed, installed_at FROM skills ORDER BY directory ASC, app_type ASC")
.map_err(|e| AppError::Database(e.to_string()))?;
let skill_iter = stmt
.query_map([], |row| {
let directory: String = row.get(0)?;
let app_type: String = row.get(1)?;
let installed: bool = row.get(2)?;
let installed_at_ts: i64 = row.get(3)?;
let installed_at =
chrono::DateTime::from_timestamp(installed_at_ts, 0).unwrap_or_default();
// 构建复合 key"app_type:directory"
let key = format!("{app_type}:{directory}");
Ok((
key,
SkillState {
installed,
installed_at,
},
))
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut skills = IndexMap::new();
for skill_res in skill_iter {
let (key, skill) = skill_res.map_err(|e| AppError::Database(e.to_string()))?;
skills.insert(key, skill);
}
Ok(skills)
}
/// 更新 Skill 状态
/// key 格式为 "app_type:directory"
pub fn update_skill_state(&self, key: &str, state: &SkillState) -> Result<(), AppError> {
// 解析 key
let (app_type, directory) = if let Some(idx) = key.find(':') {
let (app, dir) = key.split_at(idx);
(app, &dir[1..]) // 跳过冒号
} else {
// 向后兼容:如果没有前缀,默认为 claude
("claude", key)
};
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO skills (directory, app_type, installed, installed_at) VALUES (?1, ?2, ?3, ?4)",
params![directory, app_type, state.installed, state.installed_at.timestamp()],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 获取所有 Skill 仓库
pub fn get_skill_repos(&self) -> Result<Vec<SkillRepo>, AppError> {
let conn = lock_conn!(self.conn);
let mut stmt = conn
.prepare(
"SELECT owner, name, branch, enabled FROM skill_repos ORDER BY owner ASC, name ASC",
)
.map_err(|e| AppError::Database(e.to_string()))?;
let repo_iter = stmt
.query_map([], |row| {
Ok(SkillRepo {
owner: row.get(0)?,
name: row.get(1)?,
branch: row.get(2)?,
enabled: row.get(3)?,
})
})
.map_err(|e| AppError::Database(e.to_string()))?;
let mut repos = Vec::new();
for repo_res in repo_iter {
repos.push(repo_res.map_err(|e| AppError::Database(e.to_string()))?);
}
Ok(repos)
}
/// 保存 Skill 仓库
pub fn save_skill_repo(&self, repo: &SkillRepo) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"INSERT OR REPLACE INTO skill_repos (owner, name, branch, enabled) VALUES (?1, ?2, ?3, ?4)",
params![repo.owner, repo.name, repo.branch, repo.enabled],
).map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 删除 Skill 仓库
pub fn delete_skill_repo(&self, owner: &str, name: &str) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
conn.execute(
"DELETE FROM skill_repos WHERE owner = ?1 AND name = ?2",
params![owner, name],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 初始化默认的 Skill 仓库(首次启动时调用)
pub fn init_default_skill_repos(&self) -> Result<usize, AppError> {
// 检查是否已有仓库
let existing = self.get_skill_repos()?;
if !existing.is_empty() {
return Ok(0);
}
// 获取默认仓库列表
let default_store = crate::services::skill::SkillStore::default();
let mut count = 0;
for repo in &default_store.repos {
self.save_skill_repo(repo)?;
count += 1;
}
log::info!("初始化默认 Skill 仓库完成,共 {count} 个");
Ok(count)
}
}
+242
View File
@@ -0,0 +1,242 @@
//! JSON → SQLite 数据迁移
//!
//! 将旧版 config.json (MultiAppConfig) 数据迁移到 SQLite 数据库。
use super::{lock_conn, to_json_string, Database};
use crate::app_config::MultiAppConfig;
use crate::error::AppError;
use rusqlite::{params, Connection};
impl Database {
/// 从 MultiAppConfig 迁移数据到数据库
pub fn migrate_from_json(&self, config: &MultiAppConfig) -> Result<(), AppError> {
let mut conn = lock_conn!(self.conn);
let tx = conn
.transaction()
.map_err(|e| AppError::Database(e.to_string()))?;
Self::migrate_from_json_tx(&tx, config)?;
tx.commit()
.map_err(|e| AppError::Database(format!("Commit migration failed: {e}")))?;
Ok(())
}
/// 运行迁移的 dry-run 模式(在内存数据库中验证,不写入磁盘)
///
/// 用于部署前验证迁移逻辑是否正确。
pub fn migrate_from_json_dry_run(config: &MultiAppConfig) -> Result<(), AppError> {
let mut conn =
Connection::open_in_memory().map_err(|e| AppError::Database(e.to_string()))?;
Self::create_tables_on_conn(&conn)?;
Self::apply_schema_migrations_on_conn(&conn)?;
let tx = conn
.transaction()
.map_err(|e| AppError::Database(e.to_string()))?;
Self::migrate_from_json_tx(&tx, config)?;
// 显式 drop transaction 而不提交(内存数据库会被丢弃)
drop(tx);
Ok(())
}
/// 在事务中执行迁移
fn migrate_from_json_tx(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
// 1. 迁移 Providers
Self::migrate_providers(tx, config)?;
// 2. 迁移 MCP Servers
Self::migrate_mcp_servers(tx, config)?;
// 3. 迁移 Prompts
Self::migrate_prompts(tx, config)?;
// 4. 迁移 Skills
Self::migrate_skills(tx, config)?;
// 5. 迁移 Common Config
Self::migrate_common_config(tx, config)?;
Ok(())
}
/// 迁移供应商数据
fn migrate_providers(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
for (app_key, manager) in &config.apps {
let app_type = app_key;
let current_id = &manager.current;
for (id, provider) in &manager.providers {
let is_current = if id == current_id { 1 } else { 0 };
// 处理 meta 和 endpoints
let mut meta_clone = provider.meta.clone().unwrap_or_default();
let endpoints = std::mem::take(&mut meta_clone.custom_endpoints);
tx.execute(
"INSERT OR REPLACE INTO providers (
id, app_type, name, settings_config, website_url, category,
created_at, sort_index, notes, icon, icon_color, meta, is_current
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9, ?10, ?11, ?12, ?13)",
params![
id,
app_type,
provider.name,
to_json_string(&provider.settings_config)?,
provider.website_url,
provider.category,
provider.created_at,
provider.sort_index,
provider.notes,
provider.icon,
provider.icon_color,
to_json_string(&meta_clone)?,
is_current,
],
)
.map_err(|e| AppError::Database(format!("Migrate provider failed: {e}")))?;
// 迁移 Endpoints
for (url, endpoint) in endpoints {
tx.execute(
"INSERT INTO provider_endpoints (provider_id, app_type, url, added_at)
VALUES (?1, ?2, ?3, ?4)",
params![id, app_type, url, endpoint.added_at],
)
.map_err(|e| AppError::Database(format!("Migrate endpoint failed: {e}")))?;
}
}
}
Ok(())
}
/// 迁移 MCP 服务器数据
fn migrate_mcp_servers(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
if let Some(servers) = &config.mcp.servers {
for (id, server) in servers {
tx.execute(
"INSERT OR REPLACE INTO mcp_servers (
id, name, server_config, description, homepage, docs, tags,
enabled_claude, enabled_codex, enabled_gemini
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9, ?10)",
params![
id,
server.name,
to_json_string(&server.server)?,
server.description,
server.homepage,
server.docs,
to_json_string(&server.tags)?,
server.apps.claude,
server.apps.codex,
server.apps.gemini,
],
)
.map_err(|e| AppError::Database(format!("Migrate mcp server failed: {e}")))?;
}
}
Ok(())
}
/// 迁移提示词数据
fn migrate_prompts(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
let migrate_app_prompts = |prompts_map: &std::collections::HashMap<
String,
crate::prompt::Prompt,
>,
app_type: &str|
-> Result<(), AppError> {
for (id, prompt) in prompts_map {
tx.execute(
"INSERT OR REPLACE INTO prompts (
id, app_type, name, content, description, enabled, created_at, updated_at
) VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8)",
params![
id,
app_type,
prompt.name,
prompt.content,
prompt.description,
prompt.enabled,
prompt.created_at,
prompt.updated_at,
],
)
.map_err(|e| AppError::Database(format!("Migrate prompt failed: {e}")))?;
}
Ok(())
};
migrate_app_prompts(&config.prompts.claude.prompts, "claude")?;
migrate_app_prompts(&config.prompts.codex.prompts, "codex")?;
migrate_app_prompts(&config.prompts.gemini.prompts, "gemini")?;
Ok(())
}
/// 迁移 Skills 数据
fn migrate_skills(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
for (key, state) in &config.skills.skills {
tx.execute(
"INSERT OR REPLACE INTO skills (key, installed, installed_at) VALUES (?1, ?2, ?3)",
params![key, state.installed, state.installed_at.timestamp()],
)
.map_err(|e| AppError::Database(format!("Migrate skill failed: {e}")))?;
}
for repo in &config.skills.repos {
tx.execute(
"INSERT OR REPLACE INTO skill_repos (owner, name, branch, enabled) VALUES (?1, ?2, ?3, ?4)",
params![repo.owner, repo.name, repo.branch, repo.enabled],
).map_err(|e| AppError::Database(format!("Migrate skill repo failed: {e}")))?;
}
Ok(())
}
/// 迁移通用配置片段
fn migrate_common_config(
tx: &rusqlite::Transaction<'_>,
config: &MultiAppConfig,
) -> Result<(), AppError> {
if let Some(snippet) = &config.common_config_snippets.claude {
tx.execute(
"INSERT OR REPLACE INTO settings (key, value) VALUES (?1, ?2)",
params!["common_config_claude", snippet],
)
.map_err(|e| AppError::Database(format!("Migrate settings failed: {e}")))?;
}
if let Some(snippet) = &config.common_config_snippets.codex {
tx.execute(
"INSERT OR REPLACE INTO settings (key, value) VALUES (?1, ?2)",
params!["common_config_codex", snippet],
)
.map_err(|e| AppError::Database(format!("Migrate settings failed: {e}")))?;
}
if let Some(snippet) = &config.common_config_snippets.gemini {
tx.execute(
"INSERT OR REPLACE INTO settings (key, value) VALUES (?1, ?2)",
params!["common_config_gemini", snippet],
)
.map_err(|e| AppError::Database(format!("Migrate settings failed: {e}")))?;
}
Ok(())
}
}
+137
View File
@@ -0,0 +1,137 @@
//! 数据库模块 - SQLite 数据持久化
//!
//! 此模块提供应用的核心数据存储功能,包括:
//! - 供应商配置管理
//! - MCP 服务器配置
//! - 提示词管理
//! - Skills 管理
//! - 通用设置存储
//!
//! ## 架构设计
//!
//! ```text
//! database/
//! ├── mod.rs - Database 结构体 + 初始化
//! ├── schema.rs - 表结构定义 + Schema 迁移
//! ├── backup.rs - SQL 导入导出 + 快照备份
//! ├── migration.rs - JSON → SQLite 数据迁移
//! └── dao/ - 数据访问对象
//! ├── providers.rs
//! ├── mcp.rs
//! ├── prompts.rs
//! ├── skills.rs
//! └── settings.rs
//! ```
mod backup;
mod dao;
mod migration;
mod schema;
#[cfg(test)]
mod tests;
use crate::config::get_app_config_dir;
use crate::error::AppError;
use rusqlite::Connection;
use serde::Serialize;
use std::sync::Mutex;
// DAO 方法通过 impl Database 提供,无需额外导出
/// 数据库备份保留数量
const DB_BACKUP_RETAIN: usize = 10;
/// 当前 Schema 版本号
/// 每次修改表结构时递增,并在 schema.rs 中添加相应的迁移逻辑
pub(crate) const SCHEMA_VERSION: i32 = 2;
/// 安全地序列化 JSON,避免 unwrap panic
pub(crate) fn to_json_string<T: Serialize>(value: &T) -> Result<String, AppError> {
serde_json::to_string(value)
.map_err(|e| AppError::Config(format!("JSON serialization failed: {e}")))
}
/// 安全地获取 Mutex 锁,避免 unwrap panic
macro_rules! lock_conn {
($mutex:expr) => {
$mutex
.lock()
.map_err(|e| AppError::Database(format!("Mutex lock failed: {}", e)))?
};
}
// 导出宏供子模块使用
pub(crate) use lock_conn;
/// 数据库连接封装
///
/// 使用 Mutex 包装 Connection 以支持在多线程环境(如 Tauri State)中共享。
/// rusqlite::Connection 本身不是 Sync 的,因此需要这层包装。
pub struct Database {
pub(crate) conn: Mutex<Connection>,
}
impl Database {
/// 初始化数据库连接并创建表
///
/// 数据库文件位于 `~/.cc-switch/cc-switch.db`
pub fn init() -> Result<Self, AppError> {
let db_path = get_app_config_dir().join("cc-switch.db");
// 确保父目录存在
if let Some(parent) = db_path.parent() {
std::fs::create_dir_all(parent).map_err(|e| AppError::io(parent, e))?;
}
let conn = Connection::open(&db_path).map_err(|e| AppError::Database(e.to_string()))?;
// 启用外键约束
conn.execute("PRAGMA foreign_keys = ON;", [])
.map_err(|e| AppError::Database(e.to_string()))?;
let db = Self {
conn: Mutex::new(conn),
};
db.create_tables()?;
db.apply_schema_migrations()?;
db.ensure_model_pricing_seeded()?;
Ok(db)
}
/// 创建内存数据库(用于测试)
pub fn memory() -> Result<Self, AppError> {
let conn = Connection::open_in_memory().map_err(|e| AppError::Database(e.to_string()))?;
// 启用外键约束
conn.execute("PRAGMA foreign_keys = ON;", [])
.map_err(|e| AppError::Database(e.to_string()))?;
let db = Self {
conn: Mutex::new(conn),
};
db.create_tables()?;
db.ensure_model_pricing_seeded()?;
Ok(db)
}
/// 检查 MCP 服务器表是否为空
pub fn is_mcp_table_empty(&self) -> Result<bool, AppError> {
let conn = lock_conn!(self.conn);
let count: i64 = conn
.query_row("SELECT COUNT(*) FROM mcp_servers", [], |row| row.get(0))
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(count == 0)
}
/// 检查提示词表是否为空
pub fn is_prompts_table_empty(&self) -> Result<bool, AppError> {
let conn = lock_conn!(self.conn);
let count: i64 = conn
.query_row("SELECT COUNT(*) FROM prompts", [], |row| row.get(0))
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(count == 0)
}
}
+884
View File
@@ -0,0 +1,884 @@
//! Schema 定义和迁移
//!
//! 负责数据库表结构的创建和版本迁移。
use super::{lock_conn, Database, SCHEMA_VERSION};
use crate::error::AppError;
use rusqlite::Connection;
impl Database {
/// 创建所有数据库表
pub(crate) fn create_tables(&self) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
Self::create_tables_on_conn(&conn)
}
/// 在指定连接上创建表(供迁移和测试使用)
pub(crate) fn create_tables_on_conn(conn: &Connection) -> Result<(), AppError> {
// 1. Providers 表
conn.execute(
"CREATE TABLE IF NOT EXISTS providers (
id TEXT NOT NULL,
app_type TEXT NOT NULL,
name TEXT NOT NULL,
settings_config TEXT NOT NULL,
website_url TEXT,
category TEXT,
created_at INTEGER,
sort_index INTEGER,
notes TEXT,
icon TEXT,
icon_color TEXT,
meta TEXT NOT NULL DEFAULT '{}',
is_current BOOLEAN NOT NULL DEFAULT 0,
is_proxy_target BOOLEAN NOT NULL DEFAULT 0,
PRIMARY KEY (id, app_type)
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 尝试添加 is_proxy_target 列(如果表已存在但缺少该列)
let _ = conn.execute(
"ALTER TABLE providers ADD COLUMN is_proxy_target BOOLEAN NOT NULL DEFAULT 0",
[],
);
// 2. Provider Endpoints 表
conn.execute(
"CREATE TABLE IF NOT EXISTS provider_endpoints (
id INTEGER PRIMARY KEY AUTOINCREMENT,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
url TEXT NOT NULL,
added_at INTEGER,
FOREIGN KEY (provider_id, app_type) REFERENCES providers(id, app_type) ON DELETE CASCADE
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 3. MCP Servers 表
conn.execute(
"CREATE TABLE IF NOT EXISTS mcp_servers (
id TEXT PRIMARY KEY,
name TEXT NOT NULL,
server_config TEXT NOT NULL,
description TEXT,
homepage TEXT,
docs TEXT,
tags TEXT NOT NULL DEFAULT '[]',
enabled_claude BOOLEAN NOT NULL DEFAULT 0,
enabled_codex BOOLEAN NOT NULL DEFAULT 0,
enabled_gemini BOOLEAN NOT NULL DEFAULT 0
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 4. Prompts 表
conn.execute(
"CREATE TABLE IF NOT EXISTS prompts (
id TEXT NOT NULL,
app_type TEXT NOT NULL,
name TEXT NOT NULL,
content TEXT NOT NULL,
description TEXT,
enabled BOOLEAN NOT NULL DEFAULT 1,
created_at INTEGER,
updated_at INTEGER,
PRIMARY KEY (id, app_type)
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 5. Skills 表
conn.execute(
"CREATE TABLE IF NOT EXISTS skills (
directory TEXT NOT NULL,
app_type TEXT NOT NULL,
installed BOOLEAN NOT NULL DEFAULT 0,
installed_at INTEGER NOT NULL DEFAULT 0,
PRIMARY KEY (directory, app_type)
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 6. Skill Repos 表
conn.execute(
"CREATE TABLE IF NOT EXISTS skill_repos (
owner TEXT NOT NULL,
name TEXT NOT NULL,
branch TEXT NOT NULL DEFAULT 'main',
enabled BOOLEAN NOT NULL DEFAULT 1,
PRIMARY KEY (owner, name)
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 7. Settings 表 (通用配置)
conn.execute(
"CREATE TABLE IF NOT EXISTS settings (
key TEXT PRIMARY KEY,
value TEXT
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 8. Proxy Config 表 (代理服务器配置)
// 代理配置表(单例)
conn.execute(
"CREATE TABLE IF NOT EXISTS proxy_config (
id INTEGER PRIMARY KEY CHECK (id = 1),
enabled INTEGER NOT NULL DEFAULT 0,
listen_address TEXT NOT NULL DEFAULT '127.0.0.1',
listen_port INTEGER NOT NULL DEFAULT 5000,
max_retries INTEGER NOT NULL DEFAULT 3,
request_timeout INTEGER NOT NULL DEFAULT 300,
enable_logging INTEGER NOT NULL DEFAULT 1,
target_app TEXT NOT NULL DEFAULT 'claude',
created_at TEXT NOT NULL DEFAULT (datetime('now')),
updated_at TEXT NOT NULL DEFAULT (datetime('now'))
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 尝试添加 target_app 列(如果表已存在但缺少该列)
// 忽略 "duplicate column name" 错误
let _ = conn.execute(
"ALTER TABLE proxy_config ADD COLUMN target_app TEXT NOT NULL DEFAULT 'claude'",
[],
);
// 9. Provider Health 表 (Provider健康状态)
conn.execute(
"CREATE TABLE IF NOT EXISTS provider_health (
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
is_healthy INTEGER NOT NULL DEFAULT 1,
consecutive_failures INTEGER NOT NULL DEFAULT 0,
last_success_at TEXT,
last_failure_at TEXT,
last_error TEXT,
updated_at TEXT NOT NULL,
PRIMARY KEY (provider_id, app_type),
FOREIGN KEY (provider_id, app_type) REFERENCES providers(id, app_type) ON DELETE CASCADE
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 10. Proxy Usage 表 (代理使用统计,可选)
conn.execute(
"CREATE TABLE IF NOT EXISTS proxy_usage (
id INTEGER PRIMARY KEY AUTOINCREMENT,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
endpoint TEXT NOT NULL,
request_tokens INTEGER,
response_tokens INTEGER,
status_code INTEGER NOT NULL,
latency_ms INTEGER NOT NULL,
error TEXT,
timestamp TEXT NOT NULL
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 为 proxy_usage 创建索引
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_proxy_usage_timestamp
ON proxy_usage(timestamp)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_proxy_usage_provider
ON proxy_usage(provider_id, app_type)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 11. Proxy Request Logs 表 (详细请求日志)
conn.execute(
"CREATE TABLE IF NOT EXISTS proxy_request_logs (
request_id TEXT PRIMARY KEY,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
model TEXT NOT NULL,
input_tokens INTEGER NOT NULL DEFAULT 0,
output_tokens INTEGER NOT NULL DEFAULT 0,
cache_read_tokens INTEGER NOT NULL DEFAULT 0,
cache_creation_tokens INTEGER NOT NULL DEFAULT 0,
input_cost_usd TEXT NOT NULL DEFAULT '0',
output_cost_usd TEXT NOT NULL DEFAULT '0',
cache_read_cost_usd TEXT NOT NULL DEFAULT '0',
cache_creation_cost_usd TEXT NOT NULL DEFAULT '0',
total_cost_usd TEXT NOT NULL DEFAULT '0',
latency_ms INTEGER NOT NULL,
first_token_ms INTEGER,
duration_ms INTEGER,
status_code INTEGER NOT NULL,
error_message TEXT,
session_id TEXT,
provider_type TEXT,
is_streaming INTEGER NOT NULL DEFAULT 0,
cost_multiplier TEXT NOT NULL DEFAULT '1.0',
created_at INTEGER NOT NULL
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_request_logs_provider
ON proxy_request_logs(provider_id, app_type)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_request_logs_created_at
ON proxy_request_logs(created_at)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_request_logs_model
ON proxy_request_logs(model)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_request_logs_session
ON proxy_request_logs(session_id)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_request_logs_status
ON proxy_request_logs(status_code)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 12. Model Pricing 表 (模型定价)
conn.execute(
"CREATE TABLE IF NOT EXISTS model_pricing (
model_id TEXT PRIMARY KEY,
display_name TEXT NOT NULL,
input_cost_per_million TEXT NOT NULL,
output_cost_per_million TEXT NOT NULL,
cache_read_cost_per_million TEXT NOT NULL DEFAULT '0',
cache_creation_cost_per_million TEXT NOT NULL DEFAULT '0'
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 13. Usage Daily Stats 表 (每日聚合统计)
conn.execute(
"CREATE TABLE IF NOT EXISTS usage_daily_stats (
date TEXT NOT NULL,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
model TEXT NOT NULL,
request_count INTEGER NOT NULL DEFAULT 0,
total_input_tokens INTEGER NOT NULL DEFAULT 0,
total_output_tokens INTEGER NOT NULL DEFAULT 0,
total_cost_usd TEXT NOT NULL DEFAULT '0',
success_count INTEGER NOT NULL DEFAULT 0,
error_count INTEGER NOT NULL DEFAULT 0,
PRIMARY KEY (date, provider_id, app_type, model)
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
// 14. Model Test Logs 表 (模型测试日志,独立于代理使用统计)
conn.execute(
"CREATE TABLE IF NOT EXISTS model_test_logs (
id INTEGER PRIMARY KEY AUTOINCREMENT,
provider_id TEXT NOT NULL,
provider_name TEXT NOT NULL,
app_type TEXT NOT NULL,
model TEXT NOT NULL,
prompt TEXT NOT NULL,
success INTEGER NOT NULL,
message TEXT NOT NULL,
response_time_ms INTEGER,
http_status INTEGER,
tested_at INTEGER NOT NULL
)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_model_test_logs_provider
ON model_test_logs(provider_id, app_type)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
conn.execute(
"CREATE INDEX IF NOT EXISTS idx_model_test_logs_tested_at
ON model_test_logs(tested_at DESC)",
[],
)
.map_err(|e| AppError::Database(e.to_string()))?;
Ok(())
}
/// 应用 Schema 迁移
pub(crate) fn apply_schema_migrations(&self) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
Self::apply_schema_migrations_on_conn(&conn)
}
/// 在指定连接上应用 Schema 迁移
pub(crate) fn apply_schema_migrations_on_conn(conn: &Connection) -> Result<(), AppError> {
conn.execute("SAVEPOINT schema_migration;", [])
.map_err(|e| AppError::Database(format!("开启迁移 savepoint 失败: {e}")))?;
let mut version = Self::get_user_version(conn)?;
if version > SCHEMA_VERSION {
conn.execute("ROLLBACK TO schema_migration;", []).ok();
conn.execute("RELEASE schema_migration;", []).ok();
return Err(AppError::Database(format!(
"数据库版本过新({version}),当前应用仅支持 {SCHEMA_VERSION},请升级应用后再尝试。"
)));
}
let result = (|| {
while version < SCHEMA_VERSION {
match version {
0 => {
log::info!("检测到 user_version=0,迁移到 1(补齐缺失列并设置版本)");
Self::migrate_v0_to_v1(conn)?;
Self::set_user_version(conn, 1)?;
}
1 => {
log::info!(
"迁移数据库从 v1 到 v2(添加使用统计表和完整字段,重构 skills 表)"
);
Self::migrate_v1_to_v2(conn)?;
Self::set_user_version(conn, 2)?;
}
_ => {
return Err(AppError::Database(format!(
"未知的数据库版本 {version},无法迁移到 {SCHEMA_VERSION}"
)));
}
}
version = Self::get_user_version(conn)?;
}
Ok(())
})();
match result {
Ok(_) => {
conn.execute("RELEASE schema_migration;", [])
.map_err(|e| AppError::Database(format!("提交迁移 savepoint 失败: {e}")))?;
Ok(())
}
Err(e) => {
conn.execute("ROLLBACK TO schema_migration;", []).ok();
conn.execute("RELEASE schema_migration;", []).ok();
Err(e)
}
}
}
/// v0 -> v1 迁移:补齐所有缺失列
fn migrate_v0_to_v1(conn: &Connection) -> Result<(), AppError> {
// providers 表
Self::add_column_if_missing(conn, "providers", "category", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "created_at", "INTEGER")?;
Self::add_column_if_missing(conn, "providers", "sort_index", "INTEGER")?;
Self::add_column_if_missing(conn, "providers", "notes", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "icon", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "icon_color", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "meta", "TEXT NOT NULL DEFAULT '{}'")?;
Self::add_column_if_missing(
conn,
"providers",
"is_current",
"BOOLEAN NOT NULL DEFAULT 0",
)?;
// provider_endpoints 表
Self::add_column_if_missing(conn, "provider_endpoints", "added_at", "INTEGER")?;
// mcp_servers 表
Self::add_column_if_missing(conn, "mcp_servers", "description", "TEXT")?;
Self::add_column_if_missing(conn, "mcp_servers", "homepage", "TEXT")?;
Self::add_column_if_missing(conn, "mcp_servers", "docs", "TEXT")?;
Self::add_column_if_missing(conn, "mcp_servers", "tags", "TEXT NOT NULL DEFAULT '[]'")?;
Self::add_column_if_missing(
conn,
"mcp_servers",
"enabled_codex",
"BOOLEAN NOT NULL DEFAULT 0",
)?;
Self::add_column_if_missing(
conn,
"mcp_servers",
"enabled_gemini",
"BOOLEAN NOT NULL DEFAULT 0",
)?;
// prompts 表
Self::add_column_if_missing(conn, "prompts", "description", "TEXT")?;
Self::add_column_if_missing(conn, "prompts", "enabled", "BOOLEAN NOT NULL DEFAULT 1")?;
Self::add_column_if_missing(conn, "prompts", "created_at", "INTEGER")?;
Self::add_column_if_missing(conn, "prompts", "updated_at", "INTEGER")?;
// skills 表
Self::add_column_if_missing(conn, "skills", "installed_at", "INTEGER NOT NULL DEFAULT 0")?;
// skill_repos 表
Self::add_column_if_missing(
conn,
"skill_repos",
"branch",
"TEXT NOT NULL DEFAULT 'main'",
)?;
Self::add_column_if_missing(conn, "skill_repos", "enabled", "BOOLEAN NOT NULL DEFAULT 1")?;
// 注意: skills_path 字段已被移除,因为现在支持全仓库递归扫描
Ok(())
}
/// v1 -> v2 迁移:添加使用统计表和完整字段,重构 skills 表
fn migrate_v1_to_v2(conn: &Connection) -> Result<(), AppError> {
// providers 表字段
Self::add_column_if_missing(
conn,
"providers",
"cost_multiplier",
"TEXT NOT NULL DEFAULT '1.0'",
)?;
Self::add_column_if_missing(conn, "providers", "limit_daily_usd", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "limit_monthly_usd", "TEXT")?;
Self::add_column_if_missing(conn, "providers", "provider_type", "TEXT")?;
// proxy_request_logs 表(包含所有字段)
conn.execute(
"CREATE TABLE IF NOT EXISTS proxy_request_logs (
request_id TEXT PRIMARY KEY,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
model TEXT NOT NULL,
input_tokens INTEGER NOT NULL DEFAULT 0,
output_tokens INTEGER NOT NULL DEFAULT 0,
cache_read_tokens INTEGER NOT NULL DEFAULT 0,
cache_creation_tokens INTEGER NOT NULL DEFAULT 0,
input_cost_usd TEXT NOT NULL DEFAULT '0',
output_cost_usd TEXT NOT NULL DEFAULT '0',
cache_read_cost_usd TEXT NOT NULL DEFAULT '0',
cache_creation_cost_usd TEXT NOT NULL DEFAULT '0',
total_cost_usd TEXT NOT NULL DEFAULT '0',
latency_ms INTEGER NOT NULL,
first_token_ms INTEGER,
duration_ms INTEGER,
status_code INTEGER NOT NULL,
error_message TEXT,
session_id TEXT,
provider_type TEXT,
is_streaming INTEGER NOT NULL DEFAULT 0,
cost_multiplier TEXT NOT NULL DEFAULT '1.0',
created_at INTEGER NOT NULL
)",
[],
)?;
// 为已存在的表添加新字段
Self::add_column_if_missing(conn, "proxy_request_logs", "provider_type", "TEXT")?;
Self::add_column_if_missing(
conn,
"proxy_request_logs",
"is_streaming",
"INTEGER NOT NULL DEFAULT 0",
)?;
Self::add_column_if_missing(
conn,
"proxy_request_logs",
"cost_multiplier",
"TEXT NOT NULL DEFAULT '1.0'",
)?;
Self::add_column_if_missing(conn, "proxy_request_logs", "first_token_ms", "INTEGER")?;
Self::add_column_if_missing(conn, "proxy_request_logs", "duration_ms", "INTEGER")?;
// model_pricing 表
conn.execute(
"CREATE TABLE IF NOT EXISTS model_pricing (
model_id TEXT PRIMARY KEY,
display_name TEXT NOT NULL,
input_cost_per_million TEXT NOT NULL,
output_cost_per_million TEXT NOT NULL,
cache_read_cost_per_million TEXT NOT NULL DEFAULT '0',
cache_creation_cost_per_million TEXT NOT NULL DEFAULT '0'
)",
[],
)?;
// usage_daily_stats 表
conn.execute(
"CREATE TABLE IF NOT EXISTS usage_daily_stats (
date TEXT NOT NULL,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
model TEXT NOT NULL,
request_count INTEGER NOT NULL DEFAULT 0,
total_input_tokens INTEGER NOT NULL DEFAULT 0,
total_output_tokens INTEGER NOT NULL DEFAULT 0,
total_cost_usd TEXT NOT NULL DEFAULT '0',
success_count INTEGER NOT NULL DEFAULT 0,
error_count INTEGER NOT NULL DEFAULT 0,
PRIMARY KEY (date, provider_id, app_type, model)
)",
[],
)?;
// 清空并重新插入模型定价
conn.execute("DELETE FROM model_pricing", [])
.map_err(|e| AppError::Database(format!("清空模型定价失败: {e}")))?;
Self::seed_model_pricing(conn)?;
// 重构 skills 表(添加 app_type 字段)
Self::migrate_skills_table(conn)?;
Ok(())
}
/// 迁移 skills 表:从单 key 主键改为 (directory, app_type) 复合主键
fn migrate_skills_table(conn: &Connection) -> Result<(), AppError> {
// 检查是否已经是新表结构
if Self::has_column(conn, "skills", "app_type")? {
log::info!("skills 表已经包含 app_type 字段,跳过迁移");
return Ok(());
}
log::info!("开始迁移 skills 表...");
// 1. 重命名旧表
conn.execute("ALTER TABLE skills RENAME TO skills_old", [])
.map_err(|e| AppError::Database(format!("重命名旧 skills 表失败: {e}")))?;
// 2. 创建新表
conn.execute(
"CREATE TABLE skills (
directory TEXT NOT NULL,
app_type TEXT NOT NULL,
installed BOOLEAN NOT NULL DEFAULT 0,
installed_at INTEGER NOT NULL DEFAULT 0,
PRIMARY KEY (directory, app_type)
)",
[],
)
.map_err(|e| AppError::Database(format!("创建新 skills 表失败: {e}")))?;
// 3. 迁移数据:解析 key 格式(如 "claude:my-skill" 或 "codex:foo"
// 旧数据如果没有前缀,默认为 claude
let mut stmt = conn
.prepare("SELECT key, installed, installed_at FROM skills_old")
.map_err(|e| AppError::Database(format!("查询旧 skills 数据失败: {e}")))?;
let old_skills: Vec<(String, bool, i64)> = stmt
.query_map([], |row| {
Ok((
row.get::<_, String>(0)?,
row.get::<_, bool>(1)?,
row.get::<_, i64>(2)?,
))
})
.map_err(|e| AppError::Database(format!("读取旧 skills 数据失败: {e}")))?
.collect::<Result<Vec<_>, _>>()
.map_err(|e| AppError::Database(format!("解析旧 skills 数据失败: {e}")))?;
let count = old_skills.len();
for (key, installed, installed_at) in old_skills {
// 解析 key: "app:directory" 或 "directory"(默认 claude
let (app_type, directory) = if let Some(idx) = key.find(':') {
let (app, dir) = key.split_at(idx);
(app.to_string(), dir[1..].to_string()) // 跳过冒号
} else {
("claude".to_string(), key.clone())
};
conn.execute(
"INSERT INTO skills (directory, app_type, installed, installed_at) VALUES (?1, ?2, ?3, ?4)",
rusqlite::params![directory, app_type, installed, installed_at],
)
.map_err(|e| {
AppError::Database(format!("迁移 skill {key} 到新表失败: {e}"))
})?;
}
// 4. 删除旧表
conn.execute("DROP TABLE skills_old", [])
.map_err(|e| AppError::Database(format!("删除旧 skills 表失败: {e}")))?;
log::info!("skills 表迁移完成,共迁移 {count} 条记录");
Ok(())
}
/// 插入默认模型定价数据
/// 格式: (model_id, display_name, input, output, cache_read, cache_creation)
/// 注意: model_id 使用短横线格式(如 claude-haiku-4-5),与 API 返回的模型名称标准化后一致
fn seed_model_pricing(conn: &Connection) -> Result<(), AppError> {
let pricing_data = [
// Claude 4.5 系列
(
"claude-opus-4-5",
"Claude Opus 4.5",
"5",
"25",
"0.50",
"6.25",
),
(
"claude-sonnet-4-5",
"Claude Sonnet 4.5",
"3",
"15",
"0.30",
"3.75",
),
(
"claude-haiku-4-5",
"Claude Haiku 4.5",
"1",
"5",
"0.10",
"1.25",
),
// Claude 4.1 系列
(
"claude-opus-4-1",
"Claude Opus 4.1",
"15",
"75",
"1.50",
"18.75",
),
(
"claude-sonnet-4-1",
"Claude Sonnet 4.1",
"3",
"15",
"0.30",
"3.75",
),
// Claude 3.7 系列
(
"claude-sonnet-3-7",
"Claude Sonnet 3.7",
"3",
"15",
"0.30",
"3.75",
),
// Claude 3.5 系列
(
"claude-sonnet-3-5",
"Claude Sonnet 3.5",
"3",
"15",
"0.30",
"3.75",
),
(
"claude-haiku-3-5",
"Claude Haiku 3.5",
"0.80",
"4",
"0.08",
"1",
),
// GPT-5 系列(model_id 使用短横线格式)
("gpt-5", "GPT-5", "1.25", "10", "0.125", "0"),
("gpt-5-1", "GPT-5.1", "1.25", "10", "0.125", "0"),
("gpt-5-codex", "GPT-5 Codex", "1.25", "10", "0.125", "0"),
("gpt-5-1-codex", "GPT-5.1 Codex", "1.25", "10", "0.125", "0"),
// Gemini 3 系列
(
"gemini-3-pro-preview",
"Gemini 3 Pro Preview",
"2",
"12",
"0",
"0",
),
// Gemini 2.5 系列(model_id 使用短横线格式)
(
"gemini-2-5-pro",
"Gemini 2.5 Pro",
"1.25",
"10",
"0.125",
"0",
),
(
"gemini-2-5-flash",
"Gemini 2.5 Flash",
"0.3",
"2.5",
"0.03",
"0",
),
];
for (model_id, display_name, input, output, cache_read, cache_creation) in pricing_data {
conn.execute(
"INSERT OR REPLACE INTO model_pricing (
model_id, display_name, input_cost_per_million, output_cost_per_million,
cache_read_cost_per_million, cache_creation_cost_per_million
) VALUES (?1, ?2, ?3, ?4, ?5, ?6)",
rusqlite::params![
model_id,
display_name,
input,
output,
cache_read,
cache_creation
],
)
.map_err(|e| AppError::Database(format!("插入模型定价失败: {e}")))?;
}
log::info!("已插入 {} 条默认模型定价数据", pricing_data.len());
Ok(())
}
/// 确保模型定价表具备默认数据
pub fn ensure_model_pricing_seeded(&self) -> Result<(), AppError> {
let conn = lock_conn!(self.conn);
Self::ensure_model_pricing_seeded_on_conn(&conn)
}
fn ensure_model_pricing_seeded_on_conn(conn: &Connection) -> Result<(), AppError> {
let count: i64 = conn
.query_row("SELECT COUNT(*) FROM model_pricing", [], |row| row.get(0))
.map_err(|e| AppError::Database(format!("统计模型定价数据失败: {e}")))?;
if count == 0 {
Self::seed_model_pricing(conn)?;
}
Ok(())
}
// --- 辅助方法 ---
pub(crate) fn get_user_version(conn: &Connection) -> Result<i32, AppError> {
conn.query_row("PRAGMA user_version;", [], |row| row.get(0))
.map_err(|e| AppError::Database(format!("读取 user_version 失败: {e}")))
}
pub(crate) fn set_user_version(conn: &Connection, version: i32) -> Result<(), AppError> {
if version < 0 {
return Err(AppError::Database("user_version 不能为负数".to_string()));
}
let sql = format!("PRAGMA user_version = {version};");
conn.execute(&sql, [])
.map_err(|e| AppError::Database(format!("写入 user_version 失败: {e}")))?;
Ok(())
}
fn validate_identifier(s: &str, kind: &str) -> Result<(), AppError> {
if s.is_empty() {
return Err(AppError::Database(format!("{kind} 不能为空")));
}
if !s.chars().all(|c| c.is_ascii_alphanumeric() || c == '_') {
return Err(AppError::Database(format!(
"非法{kind}: {s},仅允许字母、数字和下划线"
)));
}
Ok(())
}
pub(crate) fn table_exists(conn: &Connection, table: &str) -> Result<bool, AppError> {
Self::validate_identifier(table, "表名")?;
let mut stmt = conn
.prepare("SELECT name FROM sqlite_master WHERE type='table'")
.map_err(|e| AppError::Database(format!("读取表名失败: {e}")))?;
let mut rows = stmt
.query([])
.map_err(|e| AppError::Database(format!("查询表名失败: {e}")))?;
while let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
let name: String = row
.get(0)
.map_err(|e| AppError::Database(format!("解析表名失败: {e}")))?;
if name.eq_ignore_ascii_case(table) {
return Ok(true);
}
}
Ok(false)
}
pub(crate) fn has_column(
conn: &Connection,
table: &str,
column: &str,
) -> Result<bool, AppError> {
Self::validate_identifier(table, "表名")?;
Self::validate_identifier(column, "列名")?;
let sql = format!("PRAGMA table_info(\"{table}\");");
let mut stmt = conn
.prepare(&sql)
.map_err(|e| AppError::Database(format!("读取表结构失败: {e}")))?;
let mut rows = stmt
.query([])
.map_err(|e| AppError::Database(format!("查询表结构失败: {e}")))?;
while let Some(row) = rows.next().map_err(|e| AppError::Database(e.to_string()))? {
let name: String = row
.get(1)
.map_err(|e| AppError::Database(format!("读取列名失败: {e}")))?;
if name.eq_ignore_ascii_case(column) {
return Ok(true);
}
}
Ok(false)
}
fn add_column_if_missing(
conn: &Connection,
table: &str,
column: &str,
definition: &str,
) -> Result<bool, AppError> {
Self::validate_identifier(table, "表名")?;
Self::validate_identifier(column, "列名")?;
if !Self::table_exists(conn, table)? {
return Err(AppError::Database(format!(
"{table} 不存在,无法添加列 {column}"
)));
}
if Self::has_column(conn, table, column)? {
return Ok(false);
}
let sql = format!("ALTER TABLE \"{table}\" ADD COLUMN \"{column}\" {definition};");
conn.execute(&sql, [])
.map_err(|e| AppError::Database(format!("为表 {table} 添加列 {column} 失败: {e}")))?;
log::info!("已为表 {table} 添加缺失列 {column}");
Ok(true)
}
}
+334
View File
@@ -0,0 +1,334 @@
//! 数据库模块测试
//!
//! 包含 Schema 迁移和基本功能的测试。
use super::*;
use crate::app_config::MultiAppConfig;
use crate::provider::{Provider, ProviderManager};
use indexmap::IndexMap;
use rusqlite::Connection;
use serde_json::json;
use std::collections::HashMap;
const LEGACY_SCHEMA_SQL: &str = r#"
CREATE TABLE providers (
id TEXT NOT NULL,
app_type TEXT NOT NULL,
name TEXT NOT NULL,
settings_config TEXT NOT NULL,
PRIMARY KEY (id, app_type)
);
CREATE TABLE provider_endpoints (
id INTEGER PRIMARY KEY AUTOINCREMENT,
provider_id TEXT NOT NULL,
app_type TEXT NOT NULL,
url TEXT NOT NULL
);
CREATE TABLE mcp_servers (
id TEXT PRIMARY KEY,
name TEXT NOT NULL,
server_config TEXT NOT NULL
);
CREATE TABLE prompts (
id TEXT NOT NULL,
app_type TEXT NOT NULL,
name TEXT NOT NULL,
content TEXT NOT NULL,
PRIMARY KEY (id, app_type)
);
CREATE TABLE skills (
key TEXT PRIMARY KEY,
installed BOOLEAN NOT NULL DEFAULT 0
);
CREATE TABLE skill_repos (
owner TEXT NOT NULL,
name TEXT NOT NULL,
PRIMARY KEY (owner, name)
);
CREATE TABLE settings (
key TEXT PRIMARY KEY,
value TEXT
);
"#;
#[derive(Debug)]
struct ColumnInfo {
name: String,
r#type: String,
notnull: i64,
default: Option<String>,
}
fn get_column_info(conn: &Connection, table: &str, column: &str) -> ColumnInfo {
let mut stmt = conn
.prepare(&format!("PRAGMA table_info(\"{table}\");"))
.expect("prepare pragma");
let mut rows = stmt.query([]).expect("query pragma");
while let Some(row) = rows.next().expect("read row") {
let name: String = row.get(1).expect("name");
if name.eq_ignore_ascii_case(column) {
return ColumnInfo {
name,
r#type: row.get::<_, String>(2).expect("type"),
notnull: row.get::<_, i64>(3).expect("notnull"),
default: row.get::<_, Option<String>>(4).ok().flatten(),
};
}
}
panic!("column {table}.{column} not found");
}
fn normalize_default(default: &Option<String>) -> Option<String> {
default
.as_ref()
.map(|s| s.trim_matches('\'').trim_matches('"').to_string())
}
#[test]
fn migration_sets_user_version_when_missing() {
let conn = Connection::open_in_memory().expect("open memory db");
Database::create_tables_on_conn(&conn).expect("create tables");
assert_eq!(
Database::get_user_version(&conn).expect("read version before"),
0
);
Database::apply_schema_migrations_on_conn(&conn).expect("apply migration");
assert_eq!(
Database::get_user_version(&conn).expect("read version after"),
SCHEMA_VERSION
);
}
#[test]
fn migration_rejects_future_version() {
let conn = Connection::open_in_memory().expect("open memory db");
Database::create_tables_on_conn(&conn).expect("create tables");
Database::set_user_version(&conn, SCHEMA_VERSION + 1).expect("set future version");
let err =
Database::apply_schema_migrations_on_conn(&conn).expect_err("should reject higher version");
assert!(
err.to_string().contains("数据库版本过新"),
"unexpected error: {err}"
);
}
#[test]
fn migration_adds_missing_columns_for_providers() {
let conn = Connection::open_in_memory().expect("open memory db");
// 创建旧版 providers 表,缺少新增列
conn.execute_batch(LEGACY_SCHEMA_SQL)
.expect("seed old schema");
Database::apply_schema_migrations_on_conn(&conn).expect("apply migrations");
// 验证关键新增列已补齐
for (table, column) in [
("providers", "meta"),
("providers", "is_current"),
("provider_endpoints", "added_at"),
("mcp_servers", "enabled_gemini"),
("prompts", "updated_at"),
("skills", "installed_at"),
("skill_repos", "enabled"),
] {
assert!(
Database::has_column(&conn, table, column).expect("check column"),
"{table}.{column} should exist after migration"
);
}
// 验证 meta 列约束保持一致
let meta = get_column_info(&conn, "providers", "meta");
assert_eq!(meta.notnull, 1, "meta should be NOT NULL");
assert_eq!(
normalize_default(&meta.default).as_deref(),
Some("{}"),
"meta default should be '{{}}'"
);
assert_eq!(
Database::get_user_version(&conn).expect("version after migration"),
SCHEMA_VERSION
);
}
#[test]
fn migration_aligns_column_defaults_and_types() {
let conn = Connection::open_in_memory().expect("open memory db");
conn.execute_batch(LEGACY_SCHEMA_SQL)
.expect("seed old schema");
Database::apply_schema_migrations_on_conn(&conn).expect("apply migrations");
let is_current = get_column_info(&conn, "providers", "is_current");
assert_eq!(is_current.r#type, "BOOLEAN");
assert_eq!(is_current.notnull, 1);
assert_eq!(normalize_default(&is_current.default).as_deref(), Some("0"));
let tags = get_column_info(&conn, "mcp_servers", "tags");
assert_eq!(tags.r#type, "TEXT");
assert_eq!(tags.notnull, 1);
assert_eq!(normalize_default(&tags.default).as_deref(), Some("[]"));
let enabled = get_column_info(&conn, "prompts", "enabled");
assert_eq!(enabled.r#type, "BOOLEAN");
assert_eq!(enabled.notnull, 1);
assert_eq!(normalize_default(&enabled.default).as_deref(), Some("1"));
let installed_at = get_column_info(&conn, "skills", "installed_at");
assert_eq!(installed_at.r#type, "INTEGER");
assert_eq!(installed_at.notnull, 1);
assert_eq!(
normalize_default(&installed_at.default).as_deref(),
Some("0")
);
let branch = get_column_info(&conn, "skill_repos", "branch");
assert_eq!(branch.r#type, "TEXT");
assert_eq!(normalize_default(&branch.default).as_deref(), Some("main"));
let skill_repo_enabled = get_column_info(&conn, "skill_repos", "enabled");
assert_eq!(skill_repo_enabled.r#type, "BOOLEAN");
assert_eq!(skill_repo_enabled.notnull, 1);
assert_eq!(
normalize_default(&skill_repo_enabled.default).as_deref(),
Some("1")
);
}
#[test]
fn dry_run_does_not_write_to_disk() {
// Create minimal valid config for migration
let mut apps = HashMap::new();
apps.insert("claude".to_string(), ProviderManager::default());
let config = MultiAppConfig {
version: 2,
apps,
mcp: Default::default(),
prompts: Default::default(),
skills: Default::default(),
common_config_snippets: Default::default(),
claude_common_config_snippet: None,
};
// Dry-run should succeed without any file I/O errors
let result = Database::migrate_from_json_dry_run(&config);
assert!(
result.is_ok(),
"Dry-run should succeed with valid config: {result:?}"
);
}
#[test]
fn dry_run_validates_schema_compatibility() {
// Create config with actual provider data
let mut providers = IndexMap::new();
providers.insert(
"test-provider".to_string(),
Provider {
id: "test-provider".to_string(),
name: "Test Provider".to_string(),
settings_config: json!({
"anthropicApiKey": "sk-test-123",
}),
website_url: None,
category: None,
created_at: Some(1234567890),
sort_index: None,
notes: None,
meta: None,
icon: None,
icon_color: None,
is_proxy_target: Some(false),
},
);
let mut manager = ProviderManager::default();
manager.providers = providers;
manager.current = "test-provider".to_string();
let mut apps = HashMap::new();
apps.insert("claude".to_string(), manager);
let config = MultiAppConfig {
version: 2,
apps,
mcp: Default::default(),
prompts: Default::default(),
skills: Default::default(),
common_config_snippets: Default::default(),
claude_common_config_snippet: None,
};
// Dry-run should validate the full migration path
let result = Database::migrate_from_json_dry_run(&config);
assert!(
result.is_ok(),
"Dry-run should succeed with provider data: {result:?}"
);
}
#[test]
fn model_pricing_is_seeded_on_init() {
let db = Database::memory().expect("create memory db");
let conn = db.conn.lock().expect("lock conn");
let count: i64 = conn
.query_row("SELECT COUNT(*) FROM model_pricing", [], |row| row.get(0))
.expect("count pricing");
assert!(
count > 0,
"模型定价数据应该在初始化时自动填充,实际数量: {}",
count
);
// 验证包含 Claude 模型
let claude_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM model_pricing WHERE model_id LIKE 'claude-%'",
[],
|row| row.get(0),
)
.expect("check claude");
assert!(
claude_count > 0,
"应该包含 Claude 模型定价,实际数量: {}",
claude_count
);
// 验证包含 GPT 模型
let gpt_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM model_pricing WHERE model_id LIKE 'gpt-%'",
[],
|row| row.get(0),
)
.expect("check gpt");
assert!(
gpt_count > 0,
"应该包含 GPT 模型定价,实际数量: {}",
gpt_count
);
// 验证包含 Gemini 模型
let gemini_count: i64 = conn
.query_row(
"SELECT COUNT(*) FROM model_pricing WHERE model_id LIKE 'gemini-%'",
[],
|row| row.get(0),
)
.expect("check gemini");
assert!(
gemini_count > 0,
"应该包含 Gemini 模型定价,实际数量: {}",
gemini_count
);
}
+191
View File
@@ -0,0 +1,191 @@
//! MCP server import from deep link
//!
//! Handles batch import of MCP server configurations via ccswitch:// URLs.
use super::utils::decode_base64_param;
use super::DeepLinkImportRequest;
use crate::app_config::{McpApps, McpServer};
use crate::error::AppError;
use crate::services::McpService;
use crate::store::AppState;
use serde::{Deserialize, Serialize};
use serde_json::Value;
/// MCP import result
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct McpImportResult {
/// Number of successfully imported MCP servers
pub imported_count: usize,
/// IDs of successfully imported MCP servers
pub imported_ids: Vec<String>,
/// Failed imports with error messages
pub failed: Vec<McpImportError>,
}
/// MCP import error
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct McpImportError {
/// MCP server ID
pub id: String,
/// Error message
pub error: String,
}
/// Import MCP servers from deep link request
///
/// This function handles batch import of MCP servers from standard MCP JSON format.
/// If a server already exists, only the apps flags are merged (existing config preserved).
pub fn import_mcp_from_deeplink(
state: &AppState,
request: DeepLinkImportRequest,
) -> Result<McpImportResult, AppError> {
// Verify this is an MCP request
if request.resource != "mcp" {
return Err(AppError::InvalidInput(format!(
"Expected mcp resource, got '{}'",
request.resource
)));
}
// Extract and validate apps parameter
let apps_str = request
.apps
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'apps' parameter for MCP".to_string()))?;
// Parse apps into McpApps struct
let target_apps = parse_mcp_apps(apps_str)?;
// Extract config
let config_b64 = request
.config
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'config' parameter for MCP".to_string()))?;
// Decode Base64 config
let decoded = decode_base64_param("config", config_b64)?;
let config_str = String::from_utf8(decoded)
.map_err(|e| AppError::InvalidInput(format!("Invalid UTF-8 in config: {e}")))?;
// Parse JSON
let config_json: Value = serde_json::from_str(&config_str)
.map_err(|e| AppError::InvalidInput(format!("Invalid JSON in MCP config: {e}")))?;
// Extract mcpServers object
let mcp_servers = config_json
.get("mcpServers")
.and_then(|v| v.as_object())
.ok_or_else(|| {
AppError::InvalidInput("MCP config must contain 'mcpServers' object".to_string())
})?;
if mcp_servers.is_empty() {
return Err(AppError::InvalidInput(
"No MCP servers found in config".to_string(),
));
}
// Get existing servers to check for duplicates
let existing_servers = state.db.get_all_mcp_servers()?;
// Import each MCP server
let mut imported_ids = Vec::new();
let mut failed = Vec::new();
for (id, server_spec) in mcp_servers.iter() {
// Check if server already exists
let server = if let Some(existing) = existing_servers.get(id) {
// Server exists - merge apps only, keep other fields unchanged
log::info!("MCP server '{id}' already exists, merging apps only");
let mut merged_apps = existing.apps.clone();
// Merge new apps into existing apps
if target_apps.claude {
merged_apps.claude = true;
}
if target_apps.codex {
merged_apps.codex = true;
}
if target_apps.gemini {
merged_apps.gemini = true;
}
McpServer {
id: existing.id.clone(),
name: existing.name.clone(),
server: existing.server.clone(), // Keep existing server config
apps: merged_apps, // Merged apps
description: existing.description.clone(),
homepage: existing.homepage.clone(),
docs: existing.docs.clone(),
tags: existing.tags.clone(),
}
} else {
// New server - create with provided config
log::info!("Creating new MCP server: {id}");
McpServer {
id: id.clone(),
name: id.clone(),
server: server_spec.clone(),
apps: target_apps.clone(),
description: None,
homepage: None,
docs: None,
tags: vec!["imported".to_string()],
}
};
match McpService::upsert_server(state, server) {
Ok(_) => {
imported_ids.push(id.clone());
log::info!("Successfully imported/updated MCP server: {id}");
}
Err(e) => {
failed.push(McpImportError {
id: id.clone(),
error: format!("{e}"),
});
log::warn!("Failed to import MCP server '{id}': {e}");
}
}
}
Ok(McpImportResult {
imported_count: imported_ids.len(),
imported_ids,
failed,
})
}
/// Parse apps string into McpApps struct
pub(crate) fn parse_mcp_apps(apps_str: &str) -> Result<McpApps, AppError> {
let mut apps = McpApps {
claude: false,
codex: false,
gemini: false,
};
for app in apps_str.split(',') {
match app.trim() {
"claude" => apps.claude = true,
"codex" => apps.codex = true,
"gemini" => apps.gemini = true,
other => {
return Err(AppError::InvalidInput(format!(
"Invalid app in 'apps': {other}"
)))
}
}
}
if apps.is_empty() {
return Err(AppError::InvalidInput(
"At least one app must be specified in 'apps'".to_string(),
));
}
Ok(apps)
}
+116
View File
@@ -0,0 +1,116 @@
//! Deep link import functionality for CC Switch
//!
//! This module implements the ccswitch:// protocol for importing configurations
//! via deep links. Supports importing:
//! - Provider configurations (Claude/Codex/Gemini)
//! - MCP server configurations
//! - Prompts
//! - Skills
//!
//! See docs/ccswitch-deeplink-design.md for detailed design.
mod mcp;
mod parser;
mod prompt;
mod provider;
mod skill;
mod utils;
#[cfg(test)]
mod tests;
use serde::{Deserialize, Serialize};
// Re-export public API
pub use mcp::import_mcp_from_deeplink;
pub use parser::parse_deeplink_url;
pub use prompt::import_prompt_from_deeplink;
pub use provider::{import_provider_from_deeplink, parse_and_merge_config};
pub use skill::import_skill_from_deeplink;
/// Deep link import request model
///
/// Represents a parsed ccswitch:// URL ready for processing.
/// This struct contains all possible fields for all resource types.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct DeepLinkImportRequest {
/// Protocol version (e.g., "v1")
pub version: String,
/// Resource type to import: "provider" | "prompt" | "mcp" | "skill"
pub resource: String,
// ============ Common fields ============
/// Target application (claude/codex/gemini) - for provider, prompt, skill
#[serde(skip_serializing_if = "Option::is_none")]
pub app: Option<String>,
/// Resource name
#[serde(skip_serializing_if = "Option::is_none")]
pub name: Option<String>,
/// Whether to enable after import (default: false)
#[serde(skip_serializing_if = "Option::is_none")]
pub enabled: Option<bool>,
// ============ Provider-specific fields ============
/// Provider homepage URL
#[serde(skip_serializing_if = "Option::is_none")]
pub homepage: Option<String>,
/// API endpoint/base URL
#[serde(skip_serializing_if = "Option::is_none")]
pub endpoint: Option<String>,
/// API key
#[serde(skip_serializing_if = "Option::is_none")]
pub api_key: Option<String>,
/// Optional provider icon name (maps to built-in SVG)
#[serde(skip_serializing_if = "Option::is_none")]
pub icon: Option<String>,
/// Optional model name
#[serde(skip_serializing_if = "Option::is_none")]
pub model: Option<String>,
/// Optional notes/description
#[serde(skip_serializing_if = "Option::is_none")]
pub notes: Option<String>,
/// Optional Haiku model (Claude only, v3.7.1+)
#[serde(skip_serializing_if = "Option::is_none")]
pub haiku_model: Option<String>,
/// Optional Sonnet model (Claude only, v3.7.1+)
#[serde(skip_serializing_if = "Option::is_none")]
pub sonnet_model: Option<String>,
/// Optional Opus model (Claude only, v3.7.1+)
#[serde(skip_serializing_if = "Option::is_none")]
pub opus_model: Option<String>,
// ============ Prompt-specific fields ============
/// Base64 encoded Markdown content
#[serde(skip_serializing_if = "Option::is_none")]
pub content: Option<String>,
/// Prompt description
#[serde(skip_serializing_if = "Option::is_none")]
pub description: Option<String>,
// ============ MCP-specific fields ============
/// Target applications for MCP (comma-separated: "claude,codex,gemini")
#[serde(skip_serializing_if = "Option::is_none")]
pub apps: Option<String>,
// ============ Skill-specific fields ============
/// GitHub repository (format: "owner/name")
#[serde(skip_serializing_if = "Option::is_none")]
pub repo: Option<String>,
/// Skill directory name
#[serde(skip_serializing_if = "Option::is_none")]
pub directory: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub branch: Option<String>,
// ============ Config file fields (v3.8+) ============
/// Base64 encoded config content
#[serde(skip_serializing_if = "Option::is_none")]
pub config: Option<String>,
/// Config format (json/toml)
#[serde(skip_serializing_if = "Option::is_none")]
pub config_format: Option<String>,
/// Remote config URL
#[serde(skip_serializing_if = "Option::is_none")]
pub config_url: Option<String>,
}
+313
View File
@@ -0,0 +1,313 @@
//! Deep link URL parser
//!
//! Parses ccswitch:// URLs into DeepLinkImportRequest structures.
use super::utils::validate_url;
use super::DeepLinkImportRequest;
use crate::error::AppError;
use std::collections::HashMap;
use url::Url;
/// Parse a ccswitch:// URL into a DeepLinkImportRequest
///
/// Expected format:
/// ccswitch://v1/import?resource={type}&...
pub fn parse_deeplink_url(url_str: &str) -> Result<DeepLinkImportRequest, AppError> {
// Parse URL
let url = Url::parse(url_str)
.map_err(|e| AppError::InvalidInput(format!("Invalid deep link URL: {e}")))?;
// Validate scheme
let scheme = url.scheme();
if scheme != "ccswitch" {
return Err(AppError::InvalidInput(format!(
"Invalid scheme: expected 'ccswitch', got '{scheme}'"
)));
}
// Extract version from host
let version = url
.host_str()
.ok_or_else(|| AppError::InvalidInput("Missing version in URL host".to_string()))?
.to_string();
// Validate version
if version != "v1" {
return Err(AppError::InvalidInput(format!(
"Unsupported protocol version: {version}"
)));
}
// Extract path (should be "/import")
let path = url.path();
if path != "/import" {
return Err(AppError::InvalidInput(format!(
"Invalid path: expected '/import', got '{path}'"
)));
}
// Parse query parameters
let params: HashMap<String, String> = url.query_pairs().into_owned().collect();
// Extract and validate resource type
let resource = params
.get("resource")
.ok_or_else(|| AppError::InvalidInput("Missing 'resource' parameter".to_string()))?
.clone();
// Dispatch to appropriate parser based on resource type
match resource.as_str() {
"provider" => parse_provider_deeplink(&params, version, resource),
"prompt" => parse_prompt_deeplink(&params, version, resource),
"mcp" => parse_mcp_deeplink(&params, version, resource),
"skill" => parse_skill_deeplink(&params, version, resource),
_ => Err(AppError::InvalidInput(format!(
"Unsupported resource type: {resource}"
))),
}
}
/// Parse provider deep link parameters
fn parse_provider_deeplink(
params: &HashMap<String, String>,
version: String,
resource: String,
) -> Result<DeepLinkImportRequest, AppError> {
let app = params
.get("app")
.ok_or_else(|| AppError::InvalidInput("Missing 'app' parameter".to_string()))?
.clone();
// Validate app type
if app != "claude" && app != "codex" && app != "gemini" {
return Err(AppError::InvalidInput(format!(
"Invalid app type: must be 'claude', 'codex', or 'gemini', got '{app}'"
)));
}
let name = params
.get("name")
.ok_or_else(|| AppError::InvalidInput("Missing 'name' parameter".to_string()))?
.clone();
// Make these optional for config file auto-fill (v3.8+)
let homepage = params.get("homepage").cloned();
let endpoint = params.get("endpoint").cloned();
let api_key = params.get("apiKey").cloned();
// Validate URLs only if provided
if let Some(ref hp) = homepage {
if !hp.is_empty() {
validate_url(hp, "homepage")?;
}
}
if let Some(ref ep) = endpoint {
if !ep.is_empty() {
validate_url(ep, "endpoint")?;
}
}
// Extract optional fields
let model = params.get("model").cloned();
let notes = params.get("notes").cloned();
let haiku_model = params.get("haikuModel").cloned();
let sonnet_model = params.get("sonnetModel").cloned();
let opus_model = params.get("opusModel").cloned();
let icon = params
.get("icon")
.map(|v| v.trim().to_lowercase())
.filter(|v| !v.is_empty());
let config = params.get("config").cloned();
let config_format = params.get("configFormat").cloned();
let config_url = params.get("configUrl").cloned();
let enabled = params.get("enabled").and_then(|v| v.parse::<bool>().ok());
Ok(DeepLinkImportRequest {
version,
resource,
app: Some(app),
name: Some(name),
enabled,
homepage,
endpoint,
api_key,
icon,
model,
notes,
haiku_model,
sonnet_model,
opus_model,
content: None,
description: None,
apps: None,
repo: None,
directory: None,
branch: None,
config,
config_format,
config_url,
})
}
/// Parse prompt deep link parameters
fn parse_prompt_deeplink(
params: &HashMap<String, String>,
version: String,
resource: String,
) -> Result<DeepLinkImportRequest, AppError> {
let app = params
.get("app")
.ok_or_else(|| AppError::InvalidInput("Missing 'app' parameter for prompt".to_string()))?
.clone();
// Validate app type
if app != "claude" && app != "codex" && app != "gemini" {
return Err(AppError::InvalidInput(format!(
"Invalid app type: must be 'claude', 'codex', or 'gemini', got '{app}'"
)));
}
let name = params
.get("name")
.ok_or_else(|| AppError::InvalidInput("Missing 'name' parameter for prompt".to_string()))?
.clone();
let content = params
.get("content")
.ok_or_else(|| {
AppError::InvalidInput("Missing 'content' parameter for prompt".to_string())
})?
.clone();
let description = params.get("description").cloned();
let enabled = params.get("enabled").and_then(|v| v.parse::<bool>().ok());
Ok(DeepLinkImportRequest {
version,
resource,
app: Some(app),
name: Some(name),
enabled,
content: Some(content),
description,
icon: None,
homepage: None,
endpoint: None,
api_key: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
apps: None,
repo: None,
directory: None,
branch: None,
config: None,
config_format: None,
config_url: None,
})
}
/// Parse MCP deep link parameters
fn parse_mcp_deeplink(
params: &HashMap<String, String>,
version: String,
resource: String,
) -> Result<DeepLinkImportRequest, AppError> {
let apps = params
.get("apps")
.ok_or_else(|| AppError::InvalidInput("Missing 'apps' parameter for MCP".to_string()))?
.clone();
// Validate apps format
for app in apps.split(',') {
let trimmed = app.trim();
if trimmed != "claude" && trimmed != "codex" && trimmed != "gemini" {
return Err(AppError::InvalidInput(format!(
"Invalid app in 'apps': must be 'claude', 'codex', or 'gemini', got '{trimmed}'"
)));
}
}
let config = params
.get("config")
.ok_or_else(|| AppError::InvalidInput("Missing 'config' parameter for MCP".to_string()))?
.clone();
let enabled = params.get("enabled").and_then(|v| v.parse::<bool>().ok());
Ok(DeepLinkImportRequest {
version,
resource,
apps: Some(apps),
enabled,
config: Some(config),
config_format: Some("json".to_string()), // MCP config is always JSON
app: None,
name: None,
icon: None,
homepage: None,
endpoint: None,
api_key: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
content: None,
description: None,
repo: None,
directory: None,
branch: None,
config_url: None,
})
}
/// Parse skill deep link parameters
fn parse_skill_deeplink(
params: &HashMap<String, String>,
version: String,
resource: String,
) -> Result<DeepLinkImportRequest, AppError> {
let repo = params
.get("repo")
.ok_or_else(|| AppError::InvalidInput("Missing 'repo' parameter for skill".to_string()))?
.clone();
// Validate repo format (should be "owner/name")
if !repo.contains('/') || repo.split('/').count() != 2 {
return Err(AppError::InvalidInput(format!(
"Invalid repo format: expected 'owner/name', got '{repo}'"
)));
}
let directory = params.get("directory").cloned();
let branch = params.get("branch").cloned();
Ok(DeepLinkImportRequest {
version,
resource,
repo: Some(repo),
directory,
branch,
icon: None,
app: Some("claude".to_string()), // Skills are Claude-only
name: None,
enabled: None,
homepage: None,
endpoint: None,
api_key: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
content: None,
description: None,
apps: None,
config: None,
config_format: None,
config_url: None,
})
}
+86
View File
@@ -0,0 +1,86 @@
//! Prompt import from deep link
//!
//! Handles importing prompt configurations via ccswitch:// URLs.
use super::utils::decode_base64_param;
use super::DeepLinkImportRequest;
use crate::error::AppError;
use crate::prompt::Prompt;
use crate::services::PromptService;
use crate::store::AppState;
use crate::AppType;
use std::str::FromStr;
/// Import a prompt from deep link request
pub fn import_prompt_from_deeplink(
state: &AppState,
request: DeepLinkImportRequest,
) -> Result<String, AppError> {
// Verify this is a prompt request
if request.resource != "prompt" {
return Err(AppError::InvalidInput(format!(
"Expected prompt resource, got '{}'",
request.resource
)));
}
// Extract required fields
let app_str = request
.app
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'app' field for prompt".to_string()))?;
let name = request
.name
.ok_or_else(|| AppError::InvalidInput("Missing 'name' field for prompt".to_string()))?;
// Parse app type
let app_type = AppType::from_str(app_str)
.map_err(|_| AppError::InvalidInput(format!("Invalid app type: {app_str}")))?;
// Decode content
let content_b64 = request
.content
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'content' field for prompt".to_string()))?;
let content = decode_base64_param("content", content_b64)?;
let content = String::from_utf8(content)
.map_err(|e| AppError::InvalidInput(format!("Invalid UTF-8 in content: {e}")))?;
// Generate ID
let timestamp = chrono::Utc::now().timestamp_millis();
let sanitized_name = name
.chars()
.filter(|c| c.is_alphanumeric() || *c == '-' || *c == '_')
.collect::<String>()
.to_lowercase();
let id = format!("{sanitized_name}-{timestamp}");
// Check if we should enable this prompt
let should_enable = request.enabled.unwrap_or(false);
// Create Prompt (initially disabled)
let prompt = Prompt {
id: id.clone(),
name: name.clone(),
content,
description: request.description,
enabled: false, // Always start as disabled, will be enabled later if needed
created_at: Some(timestamp),
updated_at: Some(timestamp),
};
// Save using PromptService
PromptService::upsert_prompt(state, app_type.clone(), &id, prompt)?;
// If enabled flag is set, enable this prompt (which will disable others)
if should_enable {
PromptService::enable_prompt(state, app_type, &id)?;
log::info!("Successfully imported and enabled prompt '{name}' for {app_str}");
} else {
log::info!("Successfully imported prompt '{name}' for {app_str} (disabled)");
}
Ok(id)
}
+511
View File
@@ -0,0 +1,511 @@
//! Provider import from deep link
//!
//! Handles importing provider configurations via ccswitch:// URLs.
use super::utils::{decode_base64_param, infer_homepage_from_endpoint};
use super::DeepLinkImportRequest;
use crate::error::AppError;
use crate::provider::Provider;
use crate::services::ProviderService;
use crate::store::AppState;
use crate::AppType;
use serde_json::json;
use std::str::FromStr;
/// Import a provider from a deep link request
///
/// This function:
/// 1. Validates the request
/// 2. Merges config file if provided (v3.8+)
/// 3. Converts it to a Provider structure
/// 4. Delegates to ProviderService for actual import
/// 5. Optionally sets as current provider if enabled=true
pub fn import_provider_from_deeplink(
state: &AppState,
request: DeepLinkImportRequest,
) -> Result<String, AppError> {
// Verify this is a provider request
if request.resource != "provider" {
return Err(AppError::InvalidInput(format!(
"Expected provider resource, got '{}'",
request.resource
)));
}
// Step 1: Merge config file if provided (v3.8+)
let merged_request = parse_and_merge_config(&request)?;
// Extract required fields (now as Option)
let app_str = merged_request
.app
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'app' field for provider".to_string()))?;
let api_key = merged_request.api_key.as_ref().ok_or_else(|| {
AppError::InvalidInput("API key is required (either in URL or config file)".to_string())
})?;
if api_key.is_empty() {
return Err(AppError::InvalidInput(
"API key cannot be empty".to_string(),
));
}
let endpoint = merged_request.endpoint.as_ref().ok_or_else(|| {
AppError::InvalidInput("Endpoint is required (either in URL or config file)".to_string())
})?;
if endpoint.is_empty() {
return Err(AppError::InvalidInput(
"Endpoint cannot be empty".to_string(),
));
}
let homepage = merged_request.homepage.as_ref().ok_or_else(|| {
AppError::InvalidInput("Homepage is required (either in URL or config file)".to_string())
})?;
if homepage.is_empty() {
return Err(AppError::InvalidInput(
"Homepage cannot be empty".to_string(),
));
}
let name = merged_request
.name
.as_ref()
.ok_or_else(|| AppError::InvalidInput("Missing 'name' field for provider".to_string()))?;
// Parse app type
let app_type = AppType::from_str(app_str)
.map_err(|_| AppError::InvalidInput(format!("Invalid app type: {app_str}")))?;
// Build provider configuration based on app type
let mut provider = build_provider_from_request(&app_type, &merged_request)?;
// Generate a unique ID for the provider using timestamp + sanitized name
let timestamp = chrono::Utc::now().timestamp_millis();
let sanitized_name = name
.chars()
.filter(|c| c.is_alphanumeric() || *c == '-' || *c == '_')
.collect::<String>()
.to_lowercase();
provider.id = format!("{sanitized_name}-{timestamp}");
let provider_id = provider.id.clone();
// Use ProviderService to add the provider
ProviderService::add(state, app_type.clone(), provider)?;
// If enabled=true, set as current provider
if merged_request.enabled.unwrap_or(false) {
ProviderService::switch(state, app_type.clone(), &provider_id)?;
log::info!("Provider '{provider_id}' set as current for {app_type:?}");
}
Ok(provider_id)
}
/// Build a Provider structure from a deep link request
pub(crate) fn build_provider_from_request(
app_type: &AppType,
request: &DeepLinkImportRequest,
) -> Result<Provider, AppError> {
let settings_config = match app_type {
AppType::Claude => build_claude_settings(request),
AppType::Codex => build_codex_settings(request),
AppType::Gemini => build_gemini_settings(request),
};
let provider = Provider {
id: String::new(), // Will be generated by caller
name: request.name.clone().unwrap_or_default(),
settings_config,
website_url: request.homepage.clone(),
category: None,
created_at: None,
sort_index: None,
notes: request.notes.clone(),
meta: None,
icon: request.icon.clone(),
icon_color: None,
is_proxy_target: None,
};
Ok(provider)
}
/// Build Claude settings configuration
fn build_claude_settings(request: &DeepLinkImportRequest) -> serde_json::Value {
let mut env = serde_json::Map::new();
env.insert(
"ANTHROPIC_AUTH_TOKEN".to_string(),
json!(request.api_key.clone().unwrap_or_default()),
);
env.insert(
"ANTHROPIC_BASE_URL".to_string(),
json!(request.endpoint.clone().unwrap_or_default()),
);
// Add default model if provided
if let Some(model) = &request.model {
env.insert("ANTHROPIC_MODEL".to_string(), json!(model));
}
// Add Claude-specific model fields (v3.7.1+)
if let Some(haiku_model) = &request.haiku_model {
env.insert(
"ANTHROPIC_DEFAULT_HAIKU_MODEL".to_string(),
json!(haiku_model),
);
}
if let Some(sonnet_model) = &request.sonnet_model {
env.insert(
"ANTHROPIC_DEFAULT_SONNET_MODEL".to_string(),
json!(sonnet_model),
);
}
if let Some(opus_model) = &request.opus_model {
env.insert(
"ANTHROPIC_DEFAULT_OPUS_MODEL".to_string(),
json!(opus_model),
);
}
json!({ "env": env })
}
/// Build Codex settings configuration
fn build_codex_settings(request: &DeepLinkImportRequest) -> serde_json::Value {
// Generate a safe provider name identifier
let clean_provider_name = {
let raw: String = request
.name
.clone()
.unwrap_or_else(|| "custom".to_string())
.chars()
.filter(|c| !c.is_control())
.collect();
let lower = raw.to_lowercase();
let mut key: String = lower
.chars()
.map(|c| match c {
'a'..='z' | '0'..='9' | '_' => c,
_ => '_',
})
.collect();
// Remove leading/trailing underscores
while key.starts_with('_') {
key.remove(0);
}
while key.ends_with('_') {
key.pop();
}
if key.is_empty() {
"custom".to_string()
} else {
key
}
};
// Model name: use deeplink model or default
let model_name = request
.model
.as_deref()
.unwrap_or("gpt-5-codex")
.to_string();
// Endpoint: normalize trailing slashes
let endpoint = request
.endpoint
.as_deref()
.unwrap_or("")
.trim()
.trim_end_matches('/')
.to_string();
// Build config.toml content
let config_toml = format!(
r#"model_provider = "{clean_provider_name}"
model = "{model_name}"
model_reasoning_effort = "high"
disable_response_storage = true
[model_providers.{clean_provider_name}]
name = "{clean_provider_name}"
base_url = "{endpoint}"
wire_api = "responses"
requires_openai_auth = true
"#
);
json!({
"auth": {
"OPENAI_API_KEY": request.api_key,
},
"config": config_toml
})
}
/// Build Gemini settings configuration
fn build_gemini_settings(request: &DeepLinkImportRequest) -> serde_json::Value {
let mut env = serde_json::Map::new();
env.insert("GEMINI_API_KEY".to_string(), json!(request.api_key));
env.insert(
"GOOGLE_GEMINI_BASE_URL".to_string(),
json!(request.endpoint),
);
// Add model if provided
if let Some(model) = &request.model {
env.insert("GEMINI_MODEL".to_string(), json!(model));
}
json!({ "env": env })
}
// =============================================================================
// Config Merge Logic
// =============================================================================
/// Parse and merge configuration from Base64 encoded config or remote URL
///
/// Priority: URL params > inline config > remote config
pub fn parse_and_merge_config(
request: &DeepLinkImportRequest,
) -> Result<DeepLinkImportRequest, AppError> {
// If no config provided, return original request
if request.config.is_none() && request.config_url.is_none() {
return Ok(request.clone());
}
// Step 1: Get config content
let config_content = if let Some(config_b64) = &request.config {
// Decode Base64 inline config
let decoded = decode_base64_param("config", config_b64)?;
String::from_utf8(decoded)
.map_err(|e| AppError::InvalidInput(format!("Invalid UTF-8 in config: {e}")))?
} else if let Some(_config_url) = &request.config_url {
// Fetch remote config (TODO: implement remote fetching in next phase)
return Err(AppError::InvalidInput(
"Remote config URL is not yet supported. Use inline config instead.".to_string(),
));
} else {
return Ok(request.clone());
};
// Step 2: Parse config based on format
let format = request.config_format.as_deref().unwrap_or("json");
let config_value: serde_json::Value = match format {
"json" => serde_json::from_str(&config_content)
.map_err(|e| AppError::InvalidInput(format!("Invalid JSON config: {e}")))?,
"toml" => {
let toml_value: toml::Value = toml::from_str(&config_content)
.map_err(|e| AppError::InvalidInput(format!("Invalid TOML config: {e}")))?;
// Convert TOML to JSON for uniform processing
serde_json::to_value(toml_value)
.map_err(|e| AppError::Message(format!("Failed to convert TOML to JSON: {e}")))?
}
_ => {
return Err(AppError::InvalidInput(format!(
"Unsupported config format: {format}"
)))
}
};
// Step 3: Extract values from config based on app type and merge with URL params
let mut merged = request.clone();
// MCP, Skill and other resource types don't need config merging
if request.resource != "provider" {
return Ok(merged);
}
match request.app.as_deref().unwrap_or("") {
"claude" => merge_claude_config(&mut merged, &config_value)?,
"codex" => merge_codex_config(&mut merged, &config_value)?,
"gemini" => merge_gemini_config(&mut merged, &config_value)?,
"" => {
// No app specified, skip merging
return Ok(merged);
}
_ => {
return Err(AppError::InvalidInput(format!(
"Invalid app type: {:?}",
request.app
)))
}
}
Ok(merged)
}
/// Merge Claude configuration from config file
fn merge_claude_config(
request: &mut DeepLinkImportRequest,
config: &serde_json::Value,
) -> Result<(), AppError> {
let env = config
.get("env")
.and_then(|v| v.as_object())
.ok_or_else(|| {
AppError::InvalidInput("Claude config must have 'env' object".to_string())
})?;
// Auto-fill API key if not provided in URL
if request.api_key.is_none() || request.api_key.as_ref().unwrap().is_empty() {
if let Some(token) = env.get("ANTHROPIC_AUTH_TOKEN").and_then(|v| v.as_str()) {
request.api_key = Some(token.to_string());
}
}
// Auto-fill endpoint if not provided in URL
if request.endpoint.is_none() || request.endpoint.as_ref().unwrap().is_empty() {
if let Some(base_url) = env.get("ANTHROPIC_BASE_URL").and_then(|v| v.as_str()) {
request.endpoint = Some(base_url.to_string());
}
}
// Auto-fill homepage from endpoint if not provided
if (request.homepage.is_none() || request.homepage.as_ref().unwrap().is_empty())
&& request.endpoint.is_some()
&& !request.endpoint.as_ref().unwrap().is_empty()
{
request.homepage = infer_homepage_from_endpoint(request.endpoint.as_ref().unwrap());
if request.homepage.is_none() {
request.homepage = Some("https://anthropic.com".to_string());
}
}
// Auto-fill model fields (URL params take priority)
if request.model.is_none() {
request.model = env
.get("ANTHROPIC_MODEL")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
}
if request.haiku_model.is_none() {
request.haiku_model = env
.get("ANTHROPIC_DEFAULT_HAIKU_MODEL")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
}
if request.sonnet_model.is_none() {
request.sonnet_model = env
.get("ANTHROPIC_DEFAULT_SONNET_MODEL")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
}
if request.opus_model.is_none() {
request.opus_model = env
.get("ANTHROPIC_DEFAULT_OPUS_MODEL")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
}
Ok(())
}
/// Merge Codex configuration from config file
fn merge_codex_config(
request: &mut DeepLinkImportRequest,
config: &serde_json::Value,
) -> Result<(), AppError> {
// Auto-fill API key from auth.OPENAI_API_KEY
if request.api_key.is_none() || request.api_key.as_ref().unwrap().is_empty() {
if let Some(api_key) = config
.get("auth")
.and_then(|v| v.get("OPENAI_API_KEY"))
.and_then(|v| v.as_str())
{
request.api_key = Some(api_key.to_string());
}
}
// Auto-fill endpoint and model from config string
if let Some(config_str) = config.get("config").and_then(|v| v.as_str()) {
// Parse TOML config string to extract base_url and model
if let Ok(toml_value) = toml::from_str::<toml::Value>(config_str) {
// Extract base_url from model_providers section
if request.endpoint.is_none() || request.endpoint.as_ref().unwrap().is_empty() {
if let Some(base_url) = extract_codex_base_url(&toml_value) {
request.endpoint = Some(base_url);
}
}
// Extract model
if request.model.is_none() {
if let Some(model) = toml_value.get("model").and_then(|v| v.as_str()) {
request.model = Some(model.to_string());
}
}
}
}
// Auto-fill homepage from endpoint
if (request.homepage.is_none() || request.homepage.as_ref().unwrap().is_empty())
&& request.endpoint.is_some()
&& !request.endpoint.as_ref().unwrap().is_empty()
{
request.homepage = infer_homepage_from_endpoint(request.endpoint.as_ref().unwrap());
if request.homepage.is_none() {
request.homepage = Some("https://openai.com".to_string());
}
}
Ok(())
}
/// Merge Gemini configuration from config file
fn merge_gemini_config(
request: &mut DeepLinkImportRequest,
config: &serde_json::Value,
) -> Result<(), AppError> {
// Gemini uses flat env structure
if request.api_key.is_none() || request.api_key.as_ref().unwrap().is_empty() {
if let Some(api_key) = config.get("GEMINI_API_KEY").and_then(|v| v.as_str()) {
request.api_key = Some(api_key.to_string());
}
}
if request.endpoint.is_none() || request.endpoint.as_ref().unwrap().is_empty() {
if let Some(base_url) = config.get("GEMINI_BASE_URL").and_then(|v| v.as_str()) {
request.endpoint = Some(base_url.to_string());
}
}
if request.model.is_none() {
request.model = config
.get("GEMINI_MODEL")
.and_then(|v| v.as_str())
.map(|s| s.to_string());
}
// Auto-fill homepage from endpoint
if (request.homepage.is_none() || request.homepage.as_ref().unwrap().is_empty())
&& request.endpoint.is_some()
&& !request.endpoint.as_ref().unwrap().is_empty()
{
request.homepage = infer_homepage_from_endpoint(request.endpoint.as_ref().unwrap());
if request.homepage.is_none() {
request.homepage = Some("https://ai.google.dev".to_string());
}
}
Ok(())
}
/// Extract base_url from Codex TOML config
fn extract_codex_base_url(toml_value: &toml::Value) -> Option<String> {
// Try to find base_url in model_providers section
if let Some(providers) = toml_value.get("model_providers").and_then(|v| v.as_table()) {
for (_key, provider) in providers.iter() {
if let Some(base_url) = provider.get("base_url").and_then(|v| v.as_str()) {
return Some(base_url.to_string());
}
}
}
None
}
+51
View File
@@ -0,0 +1,51 @@
//! Skill import from deep link
//!
//! Handles importing skill repository configurations via ccswitch:// URLs.
use super::DeepLinkImportRequest;
use crate::error::AppError;
use crate::services::skill::SkillRepo;
use crate::store::AppState;
/// Import a skill from deep link request
pub fn import_skill_from_deeplink(
state: &AppState,
request: DeepLinkImportRequest,
) -> Result<String, AppError> {
// Verify this is a skill request
if request.resource != "skill" {
return Err(AppError::InvalidInput(format!(
"Expected skill resource, got '{}'",
request.resource
)));
}
// Parse repo
let repo_str = request
.repo
.ok_or_else(|| AppError::InvalidInput("Missing 'repo' field for skill".to_string()))?;
let parts: Vec<&str> = repo_str.split('/').collect();
if parts.len() != 2 {
return Err(AppError::InvalidInput(format!(
"Invalid repo format: expected 'owner/name', got '{repo_str}'"
)));
}
let owner = parts[0].to_string();
let name = parts[1].to_string();
// Create SkillRepo
let repo = SkillRepo {
owner: owner.clone(),
name: name.clone(),
branch: request.branch.unwrap_or_else(|| "main".to_string()),
enabled: request.enabled.unwrap_or(true),
};
// Save using Database
state.db.save_skill_repo(&repo)?;
log::info!("Successfully added skill repo '{owner}/{name}'");
Ok(format!("{owner}/{name}"))
}
+378
View File
@@ -0,0 +1,378 @@
//! Deep link module tests
use super::mcp::parse_mcp_apps;
use super::parser::parse_deeplink_url;
use super::prompt::import_prompt_from_deeplink;
use super::provider::parse_and_merge_config;
use super::utils::{infer_homepage_from_endpoint, validate_url};
use super::DeepLinkImportRequest;
use crate::AppType;
use crate::{store::AppState, Database};
use base64::prelude::*;
use std::sync::Arc;
// =============================================================================
// Parser Tests
// =============================================================================
#[test]
fn test_parse_valid_claude_deeplink() {
let url = "ccswitch://v1/import?resource=provider&app=claude&name=Test%20Provider&homepage=https%3A%2F%2Fexample.com&endpoint=https%3A%2F%2Fapi.example.com&apiKey=sk-test-123&icon=claude";
let request = parse_deeplink_url(url).unwrap();
assert_eq!(request.version, "v1");
assert_eq!(request.resource, "provider");
assert_eq!(request.app, Some("claude".to_string()));
assert_eq!(request.name, Some("Test Provider".to_string()));
assert_eq!(request.homepage, Some("https://example.com".to_string()));
assert_eq!(
request.endpoint,
Some("https://api.example.com".to_string())
);
assert_eq!(request.api_key, Some("sk-test-123".to_string()));
assert_eq!(request.icon, Some("claude".to_string()));
}
#[test]
fn test_parse_deeplink_with_notes() {
let url = "ccswitch://v1/import?resource=provider&app=codex&name=Codex&homepage=https%3A%2F%2Fcodex.com&endpoint=https%3A%2F%2Fapi.codex.com&apiKey=key123&notes=Test%20notes";
let request = parse_deeplink_url(url).unwrap();
assert_eq!(request.notes, Some("Test notes".to_string()));
}
#[test]
fn test_parse_invalid_scheme() {
let url = "https://v1/import?resource=provider&app=claude&name=Test";
let result = parse_deeplink_url(url);
assert!(result.is_err());
assert!(result.unwrap_err().to_string().contains("Invalid scheme"));
}
#[test]
fn test_parse_unsupported_version() {
let url = "ccswitch://v2/import?resource=provider&app=claude&name=Test";
let result = parse_deeplink_url(url);
assert!(result.is_err());
assert!(result
.unwrap_err()
.to_string()
.contains("Unsupported protocol version"));
}
#[test]
fn test_parse_missing_required_field() {
// Name is still required even in v3.8+ (only homepage/endpoint/apiKey are optional)
let url = "ccswitch://v1/import?resource=provider&app=claude";
let result = parse_deeplink_url(url);
assert!(result.is_err());
assert!(result
.unwrap_err()
.to_string()
.contains("Missing 'name' parameter"));
}
// =============================================================================
// Utils Tests
// =============================================================================
#[test]
fn test_validate_invalid_url() {
let result = validate_url("not-a-url", "test");
assert!(result.is_err());
}
#[test]
fn test_validate_invalid_scheme() {
let result = validate_url("ftp://example.com", "test");
assert!(result.is_err());
assert!(result
.unwrap_err()
.to_string()
.contains("must be http or https"));
}
#[test]
fn test_infer_homepage() {
assert_eq!(
infer_homepage_from_endpoint("https://api.anthropic.com/v1"),
Some("https://anthropic.com".to_string())
);
assert_eq!(
infer_homepage_from_endpoint("https://api-test.company.com/v1"),
Some("https://test.company.com".to_string())
);
assert_eq!(
infer_homepage_from_endpoint("https://example.com"),
Some("https://example.com".to_string())
);
}
// =============================================================================
// Provider Tests
// =============================================================================
#[test]
fn test_build_gemini_provider_with_model() {
use super::provider::build_provider_from_request;
let request = DeepLinkImportRequest {
version: "v1".to_string(),
resource: "provider".to_string(),
app: Some("gemini".to_string()),
name: Some("Test Gemini".to_string()),
homepage: Some("https://example.com".to_string()),
endpoint: Some("https://api.example.com".to_string()),
api_key: Some("test-api-key".to_string()),
icon: None,
model: Some("gemini-2.0-flash".to_string()),
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
config: None,
config_format: None,
config_url: None,
apps: None,
repo: None,
directory: None,
branch: None,
content: None,
description: None,
enabled: None,
};
let provider = build_provider_from_request(&AppType::Gemini, &request).unwrap();
// Verify provider basic info
assert_eq!(provider.name, "Test Gemini");
assert_eq!(
provider.website_url,
Some("https://example.com".to_string())
);
// Verify settings_config structure
let env = provider.settings_config["env"].as_object().unwrap();
assert_eq!(env["GEMINI_API_KEY"], "test-api-key");
assert_eq!(env["GOOGLE_GEMINI_BASE_URL"], "https://api.example.com");
assert_eq!(env["GEMINI_MODEL"], "gemini-2.0-flash");
}
#[test]
fn test_build_gemini_provider_without_model() {
use super::provider::build_provider_from_request;
let request = DeepLinkImportRequest {
version: "v1".to_string(),
resource: "provider".to_string(),
app: Some("gemini".to_string()),
name: Some("Test Gemini".to_string()),
homepage: Some("https://example.com".to_string()),
endpoint: Some("https://api.example.com".to_string()),
api_key: Some("test-api-key".to_string()),
icon: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
config: None,
config_format: None,
config_url: None,
apps: None,
repo: None,
directory: None,
branch: None,
content: None,
description: None,
enabled: None,
};
let provider = build_provider_from_request(&AppType::Gemini, &request).unwrap();
let env = provider.settings_config["env"].as_object().unwrap();
assert_eq!(env["GEMINI_API_KEY"], "test-api-key");
assert_eq!(env["GOOGLE_GEMINI_BASE_URL"], "https://api.example.com");
// Model should not be present
assert!(env.get("GEMINI_MODEL").is_none());
}
#[test]
fn test_parse_and_merge_config_claude() {
// Prepare Base64 encoded Claude config
let config_json = r#"{"env":{"ANTHROPIC_AUTH_TOKEN":"sk-ant-xxx","ANTHROPIC_BASE_URL":"https://api.anthropic.com/v1","ANTHROPIC_MODEL":"claude-sonnet-4.5"}}"#;
let config_b64 = BASE64_STANDARD.encode(config_json.as_bytes());
let request = DeepLinkImportRequest {
version: "v1".to_string(),
resource: "provider".to_string(),
app: Some("claude".to_string()),
name: Some("Test".to_string()),
homepage: None,
endpoint: None,
api_key: None,
icon: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
config: Some(config_b64),
config_format: Some("json".to_string()),
config_url: None,
apps: None,
repo: None,
directory: None,
branch: None,
content: None,
description: None,
enabled: None,
};
let merged = parse_and_merge_config(&request).unwrap();
// Should auto-fill from config
assert_eq!(merged.api_key, Some("sk-ant-xxx".to_string()));
assert_eq!(
merged.endpoint,
Some("https://api.anthropic.com/v1".to_string())
);
assert_eq!(merged.homepage, Some("https://anthropic.com".to_string()));
assert_eq!(merged.model, Some("claude-sonnet-4.5".to_string()));
}
#[test]
fn test_parse_and_merge_config_url_override() {
let config_json = r#"{"env":{"ANTHROPIC_AUTH_TOKEN":"sk-old","ANTHROPIC_BASE_URL":"https://api.anthropic.com/v1"}}"#;
let config_b64 = BASE64_STANDARD.encode(config_json.as_bytes());
let request = DeepLinkImportRequest {
version: "v1".to_string(),
resource: "provider".to_string(),
app: Some("claude".to_string()),
name: Some("Test".to_string()),
homepage: None,
endpoint: None,
api_key: Some("sk-new".to_string()), // URL param should override
icon: None,
model: None,
notes: None,
haiku_model: None,
sonnet_model: None,
opus_model: None,
config: Some(config_b64),
config_format: Some("json".to_string()),
config_url: None,
apps: None,
repo: None,
directory: None,
branch: None,
content: None,
description: None,
enabled: None,
};
let merged = parse_and_merge_config(&request).unwrap();
// URL param should take priority
assert_eq!(merged.api_key, Some("sk-new".to_string()));
// Config file value should be used
assert_eq!(
merged.endpoint,
Some("https://api.anthropic.com/v1".to_string())
);
}
// =============================================================================
// Prompt Tests
// =============================================================================
#[test]
fn test_import_prompt_allows_space_in_base64_content() {
let url = "ccswitch://v1/import?resource=prompt&app=codex&name=PromptPlus&content=Pj4+";
let request = parse_deeplink_url(url).unwrap();
// URL decoded content may have "+" become space
assert_eq!(request.content.as_deref(), Some("Pj4 "));
let db = Arc::new(Database::memory().expect("create memory db"));
let state = AppState::new(db.clone());
let prompt_id = import_prompt_from_deeplink(&state, request.clone()).expect("import prompt");
let prompts = state.db.get_prompts("codex").expect("get prompts");
let prompt = prompts.get(&prompt_id).expect("prompt saved");
assert_eq!(prompt.content, ">>>");
assert_eq!(prompt.name, request.name.unwrap());
}
// =============================================================================
// MCP Tests
// =============================================================================
#[test]
fn test_parse_mcp_apps() {
let apps = parse_mcp_apps("claude,codex").unwrap();
assert!(apps.claude);
assert!(apps.codex);
assert!(!apps.gemini);
let apps = parse_mcp_apps("gemini").unwrap();
assert!(!apps.claude);
assert!(!apps.codex);
assert!(apps.gemini);
let err = parse_mcp_apps("invalid").unwrap_err();
assert!(err.to_string().contains("Invalid app"));
}
#[test]
fn test_parse_prompt_deeplink() {
let content = "Hello World";
let content_b64 = BASE64_STANDARD.encode(content);
let url = format!(
"ccswitch://v1/import?resource=prompt&app=claude&name=test&content={}&description=desc&enabled=true",
content_b64
);
let request = parse_deeplink_url(&url).unwrap();
assert_eq!(request.resource, "prompt");
assert_eq!(request.app.unwrap(), "claude");
assert_eq!(request.name.unwrap(), "test");
assert_eq!(request.content.unwrap(), content_b64);
assert_eq!(request.description.unwrap(), "desc");
assert_eq!(request.enabled.unwrap(), true);
}
#[test]
fn test_parse_mcp_deeplink() {
let config = r#"{"mcpServers":{"test":{"command":"echo"}}}"#;
let config_b64 = BASE64_STANDARD.encode(config);
let url = format!(
"ccswitch://v1/import?resource=mcp&apps=claude,codex&config={}&enabled=true",
config_b64
);
let request = parse_deeplink_url(&url).unwrap();
assert_eq!(request.resource, "mcp");
assert_eq!(request.apps.unwrap(), "claude,codex");
assert_eq!(request.config.unwrap(), config_b64);
assert_eq!(request.enabled.unwrap(), true);
}
#[test]
fn test_parse_skill_deeplink() {
let url = "ccswitch://v1/import?resource=skill&repo=owner/repo&directory=skills&branch=dev";
let request = parse_deeplink_url(&url).unwrap();
assert_eq!(request.resource, "skill");
assert_eq!(request.repo.unwrap(), "owner/repo");
assert_eq!(request.directory.unwrap(), "skills");
assert_eq!(request.branch.unwrap(), "dev");
}
+99
View File
@@ -0,0 +1,99 @@
//! Deep link utility functions
//!
//! Common helpers for URL validation, Base64 decoding, etc.
use crate::error::AppError;
use base64::prelude::*;
use url::Url;
/// Validate that a string is a valid HTTP(S) URL
pub fn validate_url(url_str: &str, field_name: &str) -> Result<(), AppError> {
let url = Url::parse(url_str)
.map_err(|e| AppError::InvalidInput(format!("Invalid URL for '{field_name}': {e}")))?;
let scheme = url.scheme();
if scheme != "http" && scheme != "https" {
return Err(AppError::InvalidInput(format!(
"Invalid URL scheme for '{field_name}': must be http or https, got '{scheme}'"
)));
}
Ok(())
}
/// Decode a Base64 parameter from deep link URL
///
/// This function handles common issues with Base64 in URLs:
/// - `+` being decoded as space
/// - Missing padding `=`
/// - Both standard and URL-safe Base64 variants
pub fn decode_base64_param(field: &str, raw: &str) -> Result<Vec<u8>, AppError> {
let mut candidates: Vec<String> = Vec::new();
// Keep spaces (to restore `+`), but remove newlines
let trimmed = raw.trim_matches(|c| c == '\r' || c == '\n');
// First try restoring spaces to "+"
if trimmed.contains(' ') {
let replaced = trimmed.replace(' ', "+");
if !replaced.is_empty() && !candidates.contains(&replaced) {
candidates.push(replaced);
}
}
// Original value
if !trimmed.is_empty() && !candidates.contains(&trimmed.to_string()) {
candidates.push(trimmed.to_string());
}
// Add padding variants
let existing = candidates.clone();
for candidate in existing {
let mut padded = candidate.clone();
let remainder = padded.len() % 4;
if remainder != 0 {
padded.extend(std::iter::repeat_n('=', 4 - remainder));
}
if !candidates.contains(&padded) {
candidates.push(padded);
}
}
let mut last_error: Option<String> = None;
for candidate in candidates {
for engine in [
&BASE64_STANDARD,
&BASE64_STANDARD_NO_PAD,
&BASE64_URL_SAFE,
&BASE64_URL_SAFE_NO_PAD,
] {
match engine.decode(&candidate) {
Ok(bytes) => return Ok(bytes),
Err(err) => last_error = Some(err.to_string()),
}
}
}
Err(AppError::InvalidInput(format!(
"{field} 参数 Base64 解码失败:{}。请确认链接参数已用 Base64 编码并经过 URL 转义(尤其是将 '+' 编码为 %2B,或使用 URL-safe Base64)。",
last_error.unwrap_or_else(|| "未知错误".to_string())
)))
}
/// Infer homepage URL from API endpoint
///
/// Examples:
/// - https://api.anthropic.com/v1 → https://anthropic.com
/// - https://api.openai.com/v1 → https://openai.com
/// - https://api-test.company.com/v1 → https://company.com
pub fn infer_homepage_from_endpoint(endpoint: &str) -> Option<String> {
let url = Url::parse(endpoint).ok()?;
let host = url.host_str()?;
// Remove common API prefixes
let clean_host = host
.strip_prefix("api.")
.or_else(|| host.strip_prefix("api-"))
.unwrap_or(host);
Some(format!("https://{clean_host}"))
}
+42
View File
@@ -50,6 +50,8 @@ pub enum AppError {
zh: String,
en: String,
},
#[error("数据库错误: {0}")]
Database(String),
}
impl AppError {
@@ -89,8 +91,48 @@ impl<T> From<PoisonError<T>> for AppError {
}
}
impl From<rusqlite::Error> for AppError {
fn from(err: rusqlite::Error) -> Self {
Self::Database(err.to_string())
}
}
impl From<AppError> for String {
fn from(err: AppError) -> Self {
err.to_string()
}
}
impl serde::Serialize for AppError {
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where
S: serde::Serializer,
{
serializer.serialize_str(&self.to_string())
}
}
/// 格式化为 JSON 错误字符串,前端可解析为结构化错误
pub fn format_skill_error(
code: &str,
context: &[(&str, &str)],
suggestion: Option<&str>,
) -> String {
use serde_json::json;
let mut ctx_map = serde_json::Map::new();
for (key, value) in context {
ctx_map.insert(key.to_string(), json!(value));
}
let error_obj = json!({
"code": code,
"context": ctx_map,
"suggestion": suggestion,
});
serde_json::to_string(&error_obj).unwrap_or_else(|_| {
// 如果 JSON 序列化失败,返回简单格式
format!("ERROR:{code}")
})
}
+656
View File
@@ -0,0 +1,656 @@
use crate::config::write_text_file;
use crate::error::AppError;
use serde_json::Value;
use std::collections::HashMap;
use std::fs;
use std::path::PathBuf;
/// 获取 Gemini 配置目录路径(支持设置覆盖)
pub fn get_gemini_dir() -> PathBuf {
if let Some(custom) = crate::settings::get_gemini_override_dir() {
return custom;
}
dirs::home_dir()
.expect("无法获取用户主目录")
.join(".gemini")
}
/// 获取 Gemini .env 文件路径
pub fn get_gemini_env_path() -> PathBuf {
get_gemini_dir().join(".env")
}
/// 解析 .env 文件内容为键值对
///
/// 此函数宽松地解析 .env 文件,跳过无效行。
/// 对于需要严格验证的场景,请使用 `parse_env_file_strict`。
pub fn parse_env_file(content: &str) -> HashMap<String, String> {
let mut map = HashMap::new();
for line in content.lines() {
let line = line.trim();
// 跳过空行和注释
if line.is_empty() || line.starts_with('#') {
continue;
}
// 解析 KEY=VALUE
if let Some((key, value)) = line.split_once('=') {
let key = key.trim().to_string();
let value = value.trim().to_string();
// 验证 key 是否有效(不为空,只包含字母、数字和下划线)
if !key.is_empty() && key.chars().all(|c| c.is_alphanumeric() || c == '_') {
map.insert(key, value);
}
}
}
map
}
/// 严格解析 .env 文件内容,返回详细的错误信息
///
/// 与 `parse_env_file` 不同,此函数在遇到无效行时会返回错误,
/// 包含行号和详细的错误信息。
///
/// # 错误
///
/// 返回 `AppError` 如果遇到以下情况:
/// - 行不包含 `=` 分隔符
/// - Key 为空或包含无效字符
/// - Key 不符合环境变量命名规范
///
/// # 使用场景
///
/// 此函数为未来的严格验证场景预留,当前运行时使用宽松的 `parse_env_file`。
/// 可用于:
/// - 配置导入验证
/// - CLI 工具的严格模式
/// - 配置文件错误诊断
///
/// 已有完整的测试覆盖,可直接使用。
#[allow(dead_code)]
pub fn parse_env_file_strict(content: &str) -> Result<HashMap<String, String>, AppError> {
let mut map = HashMap::new();
for (line_num, line) in content.lines().enumerate() {
let line = line.trim();
let line_number = line_num + 1; // 行号从 1 开始
// 跳过空行和注释
if line.is_empty() || line.starts_with('#') {
continue;
}
// 检查是否包含 =
if !line.contains('=') {
return Err(AppError::localized(
"gemini.env.parse_error.no_equals",
format!("Gemini .env 文件格式错误(第 {line_number} 行):缺少 '=' 分隔符\n行内容: {line}"),
format!("Invalid Gemini .env format (line {line_number}): missing '=' separator\nLine: {line}"),
));
}
// 解析 KEY=VALUE
if let Some((key, value)) = line.split_once('=') {
let key = key.trim();
let value = value.trim();
// 验证 key 不为空
if key.is_empty() {
return Err(AppError::localized(
"gemini.env.parse_error.empty_key",
format!("Gemini .env 文件格式错误(第 {line_number} 行):环境变量名不能为空\n行内容: {line}"),
format!("Invalid Gemini .env format (line {line_number}): variable name cannot be empty\nLine: {line}"),
));
}
// 验证 key 只包含字母、数字和下划线
if !key.chars().all(|c| c.is_alphanumeric() || c == '_') {
return Err(AppError::localized(
"gemini.env.parse_error.invalid_key",
format!("Gemini .env 文件格式错误(第 {line_number} 行):环境变量名只能包含字母、数字和下划线\n变量名: {key}"),
format!("Invalid Gemini .env format (line {line_number}): variable name can only contain letters, numbers, and underscores\nVariable: {key}"),
));
}
map.insert(key.to_string(), value.to_string());
}
}
Ok(map)
}
/// 将键值对序列化为 .env 格式
pub fn serialize_env_file(map: &HashMap<String, String>) -> String {
let mut lines = Vec::new();
// 按键排序以保证输出稳定
let mut keys: Vec<_> = map.keys().collect();
keys.sort();
for key in keys {
if let Some(value) = map.get(key) {
lines.push(format!("{key}={value}"));
}
}
lines.join("\n")
}
/// 读取 Gemini .env 文件
pub fn read_gemini_env() -> Result<HashMap<String, String>, AppError> {
let path = get_gemini_env_path();
if !path.exists() {
return Ok(HashMap::new());
}
let content = fs::read_to_string(&path).map_err(|e| AppError::io(&path, e))?;
Ok(parse_env_file(&content))
}
/// 写入 Gemini .env 文件(原子操作)
pub fn write_gemini_env_atomic(map: &HashMap<String, String>) -> Result<(), AppError> {
let path = get_gemini_env_path();
// 确保目录存在
if let Some(parent) = path.parent() {
fs::create_dir_all(parent).map_err(|e| AppError::io(parent, e))?;
// 设置目录权限为 700(仅所有者可读写执行)
#[cfg(unix)]
{
use std::os::unix::fs::PermissionsExt;
let mut perms = fs::metadata(parent)
.map_err(|e| AppError::io(parent, e))?
.permissions();
perms.set_mode(0o700);
fs::set_permissions(parent, perms).map_err(|e| AppError::io(parent, e))?;
}
}
let content = serialize_env_file(map);
write_text_file(&path, &content)?;
// 设置文件权限为 600(仅所有者可读写)
#[cfg(unix)]
{
use std::os::unix::fs::PermissionsExt;
let mut perms = fs::metadata(&path)
.map_err(|e| AppError::io(&path, e))?
.permissions();
perms.set_mode(0o600);
fs::set_permissions(&path, perms).map_err(|e| AppError::io(&path, e))?;
}
Ok(())
}
/// 从 .env 格式转换为 Provider.settings_config (JSON Value)
pub fn env_to_json(env_map: &HashMap<String, String>) -> Value {
let mut json_map = serde_json::Map::new();
for (key, value) in env_map {
json_map.insert(key.clone(), Value::String(value.clone()));
}
serde_json::json!({ "env": json_map })
}
/// 从 Provider.settings_config (JSON Value) 提取 .env 格式
pub fn json_to_env(settings: &Value) -> Result<HashMap<String, String>, AppError> {
let mut env_map = HashMap::new();
if let Some(env_obj) = settings.get("env").and_then(|v| v.as_object()) {
for (key, value) in env_obj {
if let Some(val_str) = value.as_str() {
env_map.insert(key.clone(), val_str.to_string());
}
}
}
Ok(env_map)
}
/// 验证 Gemini 配置的基本结构
///
/// 此函数只验证配置的基本格式,不强制要求 GEMINI_API_KEY。
/// 这允许用户先创建供应商配置,稍后再填写 API Key。
///
/// API Key 的验证会在切换供应商时进行(通过 `validate_gemini_settings_strict`)。
pub fn validate_gemini_settings(settings: &Value) -> Result<(), AppError> {
// 只验证基本结构,不强制要求 GEMINI_API_KEY
// 如果有 env 字段,验证它是一个对象
if let Some(env) = settings.get("env") {
if !env.is_object() {
return Err(AppError::localized(
"gemini.validation.invalid_env",
"Gemini 配置格式错误: env 必须是对象",
"Gemini config invalid: env must be an object",
));
}
}
// 如果有 config 字段,验证它是对象或 null
if let Some(config) = settings.get("config") {
if !(config.is_object() || config.is_null()) {
return Err(AppError::localized(
"gemini.validation.invalid_config",
"Gemini 配置格式错误: config 必须是对象",
"Gemini config invalid: config must be an object",
));
}
}
Ok(())
}
/// 严格验证 Gemini 配置(要求必需字段)
///
/// 此函数在切换供应商时使用,确保配置包含所有必需的字段。
/// 对于需要 API Key 的供应商(如 PackyCode),会验证 GEMINI_API_KEY 字段。
pub fn validate_gemini_settings_strict(settings: &Value) -> Result<(), AppError> {
// 先做基础格式验证(包含 env/config 类型)
validate_gemini_settings(settings)?;
let env_map = json_to_env(settings)?;
// 如果 env 为空,表示使用 OAuth(如 Google 官方),跳过验证
if env_map.is_empty() {
return Ok(());
}
// 如果 env 不为空,检查必需字段 GEMINI_API_KEY
if !env_map.contains_key("GEMINI_API_KEY") {
return Err(AppError::localized(
"gemini.validation.missing_api_key",
"Gemini 配置缺少必需字段: GEMINI_API_KEY",
"Gemini config missing required field: GEMINI_API_KEY",
));
}
Ok(())
}
/// 获取 Gemini settings.json 文件路径
///
/// 返回路径:`~/.gemini/settings.json`(与 `.env` 文件同级)
pub fn get_gemini_settings_path() -> PathBuf {
get_gemini_dir().join("settings.json")
}
/// 更新 Gemini 目录 settings.json 中的 security.auth.selectedType 字段
///
/// 此函数会:
/// 1. 读取现有的 settings.json(如果存在)
/// 2. 只更新 `security.auth.selectedType` 字段,保留其他所有字段
/// 3. 原子性写入文件
///
/// # 参数
/// - `selected_type`: 要设置的 selectedType 值(如 "gemini-api-key" 或 "oauth-personal"
fn update_selected_type(selected_type: &str) -> Result<(), AppError> {
let settings_path = get_gemini_settings_path();
// 确保目录存在
if let Some(parent) = settings_path.parent() {
fs::create_dir_all(parent).map_err(|e| AppError::io(parent, e))?;
}
// 读取现有的 settings.json(如果存在)
let mut settings_content = if settings_path.exists() {
let content =
fs::read_to_string(&settings_path).map_err(|e| AppError::io(&settings_path, e))?;
serde_json::from_str::<Value>(&content).unwrap_or_else(|_| serde_json::json!({}))
} else {
serde_json::json!({})
};
// 只更新 security.auth.selectedType 字段
if let Some(obj) = settings_content.as_object_mut() {
let security = obj
.entry("security")
.or_insert_with(|| serde_json::json!({}));
if let Some(security_obj) = security.as_object_mut() {
let auth = security_obj
.entry("auth")
.or_insert_with(|| serde_json::json!({}));
if let Some(auth_obj) = auth.as_object_mut() {
auth_obj.insert(
"selectedType".to_string(),
Value::String(selected_type.to_string()),
);
}
}
}
// 写入文件
crate::config::write_json_file(&settings_path, &settings_content)?;
Ok(())
}
/// 为 Packycode Gemini 供应商写入 settings.json
///
/// 设置 `~/.gemini/settings.json` 中的:
/// ```json
/// {
/// "security": {
/// "auth": {
/// "selectedType": "gemini-api-key"
/// }
/// }
/// }
/// ```
///
/// 保留文件中的其他所有字段。
pub fn write_packycode_settings() -> Result<(), AppError> {
update_selected_type("gemini-api-key")
}
/// 为 Google 官方 Gemini 供应商写入 settings.jsonOAuth 模式)
///
/// 设置 `~/.gemini/settings.json` 中的:
/// ```json
/// {
/// "security": {
/// "auth": {
/// "selectedType": "oauth-personal"
/// }
/// }
/// }
/// ```
///
/// 保留文件中的其他所有字段。
pub fn write_google_oauth_settings() -> Result<(), AppError> {
update_selected_type("oauth-personal")
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_parse_env_file() {
let content = r#"
# Comment line
GOOGLE_GEMINI_BASE_URL=https://example.com
GEMINI_API_KEY=sk-test123
GEMINI_MODEL=gemini-3-pro-preview
# Another comment
"#;
let map = parse_env_file(content);
assert_eq!(map.len(), 3);
assert_eq!(
map.get("GOOGLE_GEMINI_BASE_URL"),
Some(&"https://example.com".to_string())
);
assert_eq!(map.get("GEMINI_API_KEY"), Some(&"sk-test123".to_string()));
assert_eq!(
map.get("GEMINI_MODEL"),
Some(&"gemini-3-pro-preview".to_string())
);
}
#[test]
fn test_serialize_env_file() {
let mut map = HashMap::new();
map.insert("GEMINI_API_KEY".to_string(), "sk-test".to_string());
map.insert(
"GEMINI_MODEL".to_string(),
"gemini-3-pro-preview".to_string(),
);
let content = serialize_env_file(&map);
assert!(content.contains("GEMINI_API_KEY=sk-test"));
assert!(content.contains("GEMINI_MODEL=gemini-3-pro-preview"));
}
#[test]
fn test_env_json_conversion() {
let mut env_map = HashMap::new();
env_map.insert("GEMINI_API_KEY".to_string(), "test-key".to_string());
let json = env_to_json(&env_map);
let converted = json_to_env(&json).unwrap();
assert_eq!(
converted.get("GEMINI_API_KEY"),
Some(&"test-key".to_string())
);
}
#[test]
fn test_parse_env_file_strict_success() {
// 测试严格模式下正常解析
let content = r#"
# Comment line
GOOGLE_GEMINI_BASE_URL=https://example.com
GEMINI_API_KEY=sk-test123
GEMINI_MODEL=gemini-3-pro-preview
# Another comment
"#;
let result = parse_env_file_strict(content);
assert!(result.is_ok());
let map = result.unwrap();
assert_eq!(map.len(), 3);
assert_eq!(
map.get("GOOGLE_GEMINI_BASE_URL"),
Some(&"https://example.com".to_string())
);
assert_eq!(map.get("GEMINI_API_KEY"), Some(&"sk-test123".to_string()));
assert_eq!(
map.get("GEMINI_MODEL"),
Some(&"gemini-3-pro-preview".to_string())
);
}
#[test]
fn test_parse_env_file_strict_missing_equals() {
// 测试严格模式下检测缺少 = 的行
let content = "GOOGLE_GEMINI_BASE_URL=https://example.com
INVALID_LINE_WITHOUT_EQUALS
GEMINI_API_KEY=sk-test123";
let result = parse_env_file_strict(content);
assert!(result.is_err());
let err = result.unwrap_err();
let err_msg = format!("{err:?}");
assert!(err_msg.contains("第 2 行") || err_msg.contains("line 2"));
assert!(err_msg.contains("INVALID_LINE_WITHOUT_EQUALS"));
}
#[test]
fn test_parse_env_file_strict_empty_key() {
// 测试严格模式下检测空 key
let content = "GOOGLE_GEMINI_BASE_URL=https://example.com
=value_without_key
GEMINI_API_KEY=sk-test123";
let result = parse_env_file_strict(content);
assert!(result.is_err());
let err = result.unwrap_err();
let err_msg = format!("{err:?}");
assert!(err_msg.contains("第 2 行") || err_msg.contains("line 2"));
assert!(err_msg.contains("empty") || err_msg.contains(""));
}
#[test]
fn test_parse_env_file_strict_invalid_key_characters() {
// 测试严格模式下检测无效字符(如空格、特殊符号)
let content = "GOOGLE_GEMINI_BASE_URL=https://example.com
INVALID KEY WITH SPACES=value
GEMINI_API_KEY=sk-test123";
let result = parse_env_file_strict(content);
assert!(result.is_err());
let err = result.unwrap_err();
let err_msg = format!("{err:?}");
assert!(err_msg.contains("第 2 行") || err_msg.contains("line 2"));
assert!(err_msg.contains("INVALID KEY WITH SPACES"));
}
#[test]
fn test_parse_env_file_lax_vs_strict() {
// 测试宽松模式和严格模式的差异
let content = "VALID_KEY=value
INVALID LINE
KEY_WITH-DASH=value";
// 宽松模式:跳过无效行,继续解析
let lax_result = parse_env_file(content);
assert_eq!(lax_result.len(), 1); // 只有 VALID_KEY
assert_eq!(lax_result.get("VALID_KEY"), Some(&"value".to_string()));
// 严格模式:遇到无效行立即返回错误
let strict_result = parse_env_file_strict(content);
assert!(strict_result.is_err());
}
#[test]
fn test_packycode_settings_structure() {
// 验证 Packycode settings.json 的结构正确
let settings_content = serde_json::json!({
"security": {
"auth": {
"selectedType": "gemini-api-key"
}
}
});
assert_eq!(
settings_content["security"]["auth"]["selectedType"],
"gemini-api-key"
);
}
#[test]
fn test_packycode_settings_merge() {
// 测试合并逻辑:应该保留其他字段
let mut existing_settings = serde_json::json!({
"otherField": "should-be-kept",
"security": {
"otherSetting": "also-kept",
"auth": {
"otherAuth": "preserved"
}
}
});
// 模拟更新 selectedType
if let Some(obj) = existing_settings.as_object_mut() {
let security = obj
.entry("security")
.or_insert_with(|| serde_json::json!({}));
if let Some(security_obj) = security.as_object_mut() {
let auth = security_obj
.entry("auth")
.or_insert_with(|| serde_json::json!({}));
if let Some(auth_obj) = auth.as_object_mut() {
auth_obj.insert(
"selectedType".to_string(),
Value::String("gemini-api-key".to_string()),
);
}
}
}
// 验证所有字段都被保留
assert_eq!(existing_settings["otherField"], "should-be-kept");
assert_eq!(existing_settings["security"]["otherSetting"], "also-kept");
assert_eq!(
existing_settings["security"]["auth"]["otherAuth"],
"preserved"
);
assert_eq!(
existing_settings["security"]["auth"]["selectedType"],
"gemini-api-key"
);
}
#[test]
fn test_google_oauth_settings_structure() {
// 验证 Google OAuth settings.json 的结构正确
let settings_content = serde_json::json!({
"security": {
"auth": {
"selectedType": "oauth-personal"
}
}
});
assert_eq!(
settings_content["security"]["auth"]["selectedType"],
"oauth-personal"
);
}
#[test]
fn test_validate_empty_env_for_oauth() {
// 测试空 envGoogle 官方 OAuth)可以通过基本验证
let settings = serde_json::json!({
"env": {}
});
assert!(validate_gemini_settings(&settings).is_ok());
// 严格验证也应该通过(空 env 表示 OAuth)
assert!(validate_gemini_settings_strict(&settings).is_ok());
}
#[test]
fn test_validate_env_with_api_key() {
// 测试有 API Key 的配置可以通过验证
let settings = serde_json::json!({
"env": {
"GEMINI_API_KEY": "sk-test123",
"GEMINI_MODEL": "gemini-3-pro-preview"
}
});
assert!(validate_gemini_settings(&settings).is_ok());
assert!(validate_gemini_settings_strict(&settings).is_ok());
}
#[test]
fn test_validate_env_without_api_key_relaxed() {
// 测试缺少 API Key 的非空配置在基本验证中可以通过(用户稍后填写)
let settings = serde_json::json!({
"env": {
"GEMINI_MODEL": "gemini-3-pro-preview"
}
});
// 基本验证应该通过(允许稍后填写 API Key)
assert!(validate_gemini_settings(&settings).is_ok());
// 严格验证应该失败(切换时要求完整配置)
assert!(validate_gemini_settings_strict(&settings).is_err());
}
#[test]
fn test_validate_invalid_env_type() {
// 测试 env 不是对象时会失败
let settings = serde_json::json!({
"env": "invalid_string"
});
assert!(validate_gemini_settings(&settings).is_err());
}
}
+142
View File
@@ -0,0 +1,142 @@
use serde::{Deserialize, Serialize};
use serde_json::{Map, Value};
use std::fs;
use std::path::{Path, PathBuf};
use crate::config::atomic_write;
use crate::error::AppError;
use crate::gemini_config::get_gemini_settings_path;
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct McpStatus {
pub user_config_path: String,
pub user_config_exists: bool,
pub server_count: usize,
}
/// 获取 Gemini MCP 配置文件路径(~/.gemini/settings.json
fn user_config_path() -> PathBuf {
get_gemini_settings_path()
}
fn read_json_value(path: &Path) -> Result<Value, AppError> {
if !path.exists() {
return Ok(serde_json::json!({}));
}
let content = fs::read_to_string(path).map_err(|e| AppError::io(path, e))?;
let value: Value = serde_json::from_str(&content).map_err(|e| AppError::json(path, e))?;
Ok(value)
}
fn write_json_value(path: &Path, value: &Value) -> Result<(), AppError> {
if let Some(parent) = path.parent() {
fs::create_dir_all(parent).map_err(|e| AppError::io(parent, e))?;
}
let json =
serde_json::to_string_pretty(value).map_err(|e| AppError::JsonSerialize { source: e })?;
atomic_write(path, json.as_bytes())
}
/// 读取 Gemini settings.json 中的 mcpServers 映射
///
/// 执行反向格式转换以保持与统一 MCP 结构的兼容性:
/// - httpUrl → url + type: "http"
/// - 仅有 url 字段 → 保持不变(SSE 类型)
/// - 仅有 command 字段 → 保持不变(stdio 类型)
pub fn read_mcp_servers_map() -> Result<std::collections::HashMap<String, Value>, AppError> {
let path = user_config_path();
if !path.exists() {
return Ok(std::collections::HashMap::new());
}
let root = read_json_value(&path)?;
let mut servers: std::collections::HashMap<String, Value> = root
.get("mcpServers")
.and_then(|v| v.as_object())
.map(|obj| obj.iter().map(|(k, v)| (k.clone(), v.clone())).collect())
.unwrap_or_default();
// 反向格式转换:Gemini 特有格式 → 统一 MCP 格式
for (_, spec) in servers.iter_mut() {
if let Some(obj) = spec.as_object_mut() {
// httpUrl → url + type: "http"
if let Some(http_url) = obj.remove("httpUrl") {
obj.insert("url".to_string(), http_url);
obj.insert("type".to_string(), Value::String("http".to_string()));
}
// 如果有 url 但没有 type,不添加 type(默认为 SSE
// 如果有 command 但没有 type,不添加 type(默认为 stdio
}
}
Ok(servers)
}
/// 将给定的启用 MCP 服务器映射写入到 Gemini settings.json 的 mcpServers 字段
/// 仅覆盖 mcpServers,其他字段保持不变
pub fn set_mcp_servers_map(
servers: &std::collections::HashMap<String, Value>,
) -> Result<(), AppError> {
let path = user_config_path();
let mut root = if path.exists() {
read_json_value(&path)?
} else {
serde_json::json!({})
};
// 构建 mcpServers 对象:移除 UI 辅助字段(enabled/source),仅保留实际 MCP 规范
let mut out: Map<String, Value> = Map::new();
for (id, spec) in servers.iter() {
let mut obj = if let Some(map) = spec.as_object() {
map.clone()
} else {
return Err(AppError::McpValidation(format!(
"MCP 服务器 '{id}' 不是对象"
)));
};
// 提取 server 字段(如果存在)
if let Some(server_val) = obj.remove("server") {
let server_obj = server_val.as_object().cloned().ok_or_else(|| {
AppError::McpValidation(format!("MCP 服务器 '{id}' server 字段不是对象"))
})?;
obj = server_obj;
}
// Gemini CLI 格式转换:
// - Gemini 不使用 "type" 字段(从字段名推断传输类型)
// - HTTP 使用 "httpUrl" 字段,SSE 使用 "url" 字段
let transport_type = obj.get("type").and_then(|v| v.as_str());
if transport_type == Some("http") {
// HTTP streaming: 将 "url" 重命名为 "httpUrl"
if let Some(url_value) = obj.remove("url") {
obj.insert("httpUrl".to_string(), url_value);
}
}
// SSE 保持 "url" 字段不变
// 移除 UI 辅助字段和 type 字段(Gemini 不需要)
obj.remove("type");
obj.remove("enabled");
obj.remove("source");
obj.remove("id");
obj.remove("name");
obj.remove("description");
obj.remove("tags");
obj.remove("homepage");
obj.remove("docs");
out.insert(id.clone(), Value::Object(obj));
}
{
let obj = root
.as_object_mut()
.ok_or_else(|| AppError::Config("~/.gemini/settings.json 根必须是对象".into()))?;
obj.insert("mcpServers".into(), Value::Object(out));
}
write_json_value(&path, &root)?;
Ok(())
}
+29
View File
@@ -13,7 +13,9 @@ fn cell() -> &'static RwLock<Option<InitErrorPayload>> {
INIT_ERROR.get_or_init(|| RwLock::new(None))
}
#[allow(dead_code)]
pub fn set_init_error(payload: InitErrorPayload) {
#[allow(clippy::unwrap_used)]
if let Ok(mut guard) = cell().write() {
*guard = Some(payload);
}
@@ -23,6 +25,33 @@ pub fn get_init_error() -> Option<InitErrorPayload> {
cell().read().ok()?.clone()
}
// ============================================================
// 迁移结果状态
// ============================================================
static MIGRATION_SUCCESS: OnceLock<RwLock<bool>> = OnceLock::new();
fn migration_cell() -> &'static RwLock<bool> {
MIGRATION_SUCCESS.get_or_init(|| RwLock::new(false))
}
pub fn set_migration_success() {
if let Ok(mut guard) = migration_cell().write() {
*guard = true;
}
}
/// 获取并消费迁移成功状态(只返回一次 true,之后返回 false)
pub fn take_migration_success() -> bool {
if let Ok(mut guard) = migration_cell().write() {
let val = *guard;
*guard = false;
val
} else {
false
}
}
#[cfg(test)]
mod tests {
use super::*;
+574 -344
View File
File diff suppressed because it is too large Load Diff
+10
View File
@@ -2,5 +2,15 @@
#![cfg_attr(not(debug_assertions), windows_subsystem = "windows")]
fn main() {
// 在 Linux 上设置 WebKit 环境变量以解决 DMA-BUF 渲染问题
// 某些 Linux 系统(如 Debian 13.2、Nvidia GPU)上 WebKitGTK 的 DMA-BUF 渲染器可能导致白屏/黑屏
// 参考: https://github.com/tauri-apps/tauri/issues/9394
#[cfg(target_os = "linux")]
{
if std::env::var("WEBKIT_DISABLE_DMABUF_RENDERER").is_err() {
std::env::set_var("WEBKIT_DISABLE_DMABUF_RENDERER", "1");
}
}
cc_switch_lib::run();
}
-720
View File
@@ -1,720 +0,0 @@
use serde_json::{json, Value};
use std::collections::HashMap;
use crate::app_config::{AppType, McpConfig, MultiAppConfig};
use crate::error::AppError;
/// 基础校验:允许 stdio/http;或省略 type(视为 stdio)。对应必填字段存在
fn validate_server_spec(spec: &Value) -> Result<(), AppError> {
if !spec.is_object() {
return Err(AppError::McpValidation(
"MCP 服务器连接定义必须为 JSON 对象".into(),
));
}
let t_opt = spec.get("type").and_then(|x| x.as_str());
// 支持两种:stdio/http;若缺省 type 则按 stdio 处理(与社区常见 .mcp.json 一致)
let is_stdio = t_opt.map(|t| t == "stdio").unwrap_or(true);
let is_http = t_opt.map(|t| t == "http").unwrap_or(false);
if !(is_stdio || is_http) {
return Err(AppError::McpValidation(
"MCP 服务器 type 必须是 'stdio' 或 'http'(或省略表示 stdio".into(),
));
}
if is_stdio {
let cmd = spec.get("command").and_then(|x| x.as_str()).unwrap_or("");
if cmd.trim().is_empty() {
return Err(AppError::McpValidation(
"stdio 类型的 MCP 服务器缺少 command 字段".into(),
));
}
}
if is_http {
let url = spec.get("url").and_then(|x| x.as_str()).unwrap_or("");
if url.trim().is_empty() {
return Err(AppError::McpValidation(
"http 类型的 MCP 服务器缺少 url 字段".into(),
));
}
}
Ok(())
}
fn validate_mcp_entry(entry: &Value) -> Result<(), AppError> {
let obj = entry
.as_object()
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目必须为 JSON 对象".into()))?;
let server = obj
.get("server")
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目缺少 server 字段".into()))?;
validate_server_spec(server)?;
for key in ["name", "description", "homepage", "docs"] {
if let Some(val) = obj.get(key) {
if !val.is_string() {
return Err(AppError::McpValidation(format!(
"MCP 服务器 {} 必须为字符串",
key
)));
}
}
}
if let Some(tags) = obj.get("tags") {
let arr = tags
.as_array()
.ok_or_else(|| AppError::McpValidation("MCP 服务器 tags 必须为字符串数组".into()))?;
if !arr.iter().all(|item| item.is_string()) {
return Err(AppError::McpValidation(
"MCP 服务器 tags 必须为字符串数组".into(),
));
}
}
if let Some(enabled) = obj.get("enabled") {
if !enabled.is_boolean() {
return Err(AppError::McpValidation(
"MCP 服务器 enabled 必须为布尔值".into(),
));
}
}
Ok(())
}
fn normalize_server_keys(map: &mut HashMap<String, Value>) -> usize {
let mut change_count = 0usize;
let mut renames: Vec<(String, String)> = Vec::new();
for (key_ref, value) in map.iter_mut() {
let key = key_ref.clone();
let Some(obj) = value.as_object_mut() else {
continue;
};
let id_value = obj.get("id").cloned();
let target_id: String;
match id_value {
Some(id_val) => match id_val.as_str() {
Some(id_str) => {
let trimmed = id_str.trim();
if trimmed.is_empty() {
obj.insert("id".into(), json!(key.clone()));
change_count += 1;
target_id = key.clone();
} else {
if trimmed != id_str {
obj.insert("id".into(), json!(trimmed));
change_count += 1;
}
target_id = trimmed.to_string();
}
}
None => {
obj.insert("id".into(), json!(key.clone()));
change_count += 1;
target_id = key.clone();
}
},
None => {
obj.insert("id".into(), json!(key.clone()));
change_count += 1;
target_id = key.clone();
}
}
if target_id != key {
renames.push((key, target_id));
}
}
for (old_key, new_key) in renames {
if old_key == new_key {
continue;
}
if map.contains_key(&new_key) {
log::warn!(
"MCP 条目 '{}' 的内部 id '{}' 与现有键冲突,回退为原键",
old_key,
new_key
);
if let Some(value) = map.get_mut(&old_key) {
if let Some(obj) = value.as_object_mut() {
if obj
.get("id")
.and_then(|v| v.as_str())
.map(|s| s != old_key)
.unwrap_or(true)
{
obj.insert("id".into(), json!(old_key.clone()));
change_count += 1;
}
}
}
continue;
}
if let Some(mut value) = map.remove(&old_key) {
if let Some(obj) = value.as_object_mut() {
obj.insert("id".into(), json!(new_key.clone()));
}
log::info!("MCP 条目键名已自动修复: '{}' -> '{}'", old_key, new_key);
map.insert(new_key, value);
change_count += 1;
}
}
change_count
}
pub fn normalize_servers_for(config: &mut MultiAppConfig, app: &AppType) -> usize {
let servers = &mut config.mcp_for_mut(app).servers;
normalize_server_keys(servers)
}
fn extract_server_spec(entry: &Value) -> Result<Value, AppError> {
let obj = entry
.as_object()
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目必须为 JSON 对象".into()))?;
let server = obj
.get("server")
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目缺少 server 字段".into()))?;
if !server.is_object() {
return Err(AppError::McpValidation(
"MCP 服务器 server 字段必须为 JSON 对象".into(),
));
}
Ok(server.clone())
}
/// 返回已启用的 MCP 服务器(过滤 enabled==true
fn collect_enabled_servers(cfg: &McpConfig) -> HashMap<String, Value> {
let mut out = HashMap::new();
for (id, entry) in cfg.servers.iter() {
let enabled = entry
.get("enabled")
.and_then(|v| v.as_bool())
.unwrap_or(false);
if !enabled {
continue;
}
match extract_server_spec(entry) {
Ok(spec) => {
out.insert(id.clone(), spec);
}
Err(err) => {
log::warn!("跳过无效的 MCP 条目 '{}': {}", id, err);
}
}
}
out
}
pub fn get_servers_snapshot_for(
config: &mut MultiAppConfig,
app: &AppType,
) -> (HashMap<String, Value>, usize) {
let normalized = normalize_servers_for(config, app);
let mut snapshot = config.mcp_for(app).servers.clone();
snapshot.retain(|id, value| {
let Some(obj) = value.as_object_mut() else {
log::warn!("跳过无效的 MCP 条目 '{}': 必须为 JSON 对象", id);
return false;
};
obj.entry(String::from("id")).or_insert(json!(id));
match validate_mcp_entry(value) {
Ok(()) => true,
Err(err) => {
log::error!("config.json 中存在无效的 MCP 条目 '{}': {}", id, err);
false
}
}
});
(snapshot, normalized)
}
pub fn upsert_in_config_for(
config: &mut MultiAppConfig,
app: &AppType,
id: &str,
spec: Value,
) -> Result<bool, AppError> {
if id.trim().is_empty() {
return Err(AppError::InvalidInput("MCP 服务器 ID 不能为空".into()));
}
normalize_servers_for(config, app);
validate_mcp_entry(&spec)?;
let mut entry_obj = spec
.as_object()
.cloned()
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目必须为 JSON 对象".into()))?;
if let Some(existing_id) = entry_obj.get("id") {
let Some(existing_id_str) = existing_id.as_str() else {
return Err(AppError::McpValidation("MCP 服务器 id 必须为字符串".into()));
};
if existing_id_str != id {
return Err(AppError::McpValidation(format!(
"MCP 服务器条目中的 id '{}' 与参数 id '{}' 不一致",
existing_id_str, id
)));
}
} else {
entry_obj.insert(String::from("id"), json!(id));
}
let value = Value::Object(entry_obj);
let servers = &mut config.mcp_for_mut(app).servers;
let before = servers.get(id).cloned();
servers.insert(id.to_string(), value);
Ok(before.is_none())
}
pub fn delete_in_config_for(
config: &mut MultiAppConfig,
app: &AppType,
id: &str,
) -> Result<bool, AppError> {
if id.trim().is_empty() {
return Err(AppError::InvalidInput("MCP 服务器 ID 不能为空".into()));
}
normalize_servers_for(config, app);
let existed = config.mcp_for_mut(app).servers.remove(id).is_some();
Ok(existed)
}
/// 设置启用状态(不执行落盘或文件同步)
pub fn set_enabled_flag_for(
config: &mut MultiAppConfig,
app: &AppType,
id: &str,
enabled: bool,
) -> Result<bool, AppError> {
if id.trim().is_empty() {
return Err(AppError::InvalidInput("MCP 服务器 ID 不能为空".into()));
}
normalize_servers_for(config, app);
if let Some(spec) = config.mcp_for_mut(app).servers.get_mut(id) {
// 写入 enabled 字段
let mut obj = spec
.as_object()
.cloned()
.ok_or_else(|| AppError::McpValidation("MCP 服务器定义必须为 JSON 对象".into()))?;
obj.insert("enabled".into(), json!(enabled));
*spec = Value::Object(obj);
} else {
// 若不存在则直接返回 false
return Ok(false);
}
Ok(true)
}
/// 将 config.json 中 enabled==true 的项投影写入 ~/.claude.json
pub fn sync_enabled_to_claude(config: &MultiAppConfig) -> Result<(), AppError> {
let enabled = collect_enabled_servers(&config.mcp.claude);
crate::claude_mcp::set_mcp_servers_map(&enabled)
}
/// 从 ~/.claude.json 导入 mcpServers 到 config.json(设为 enabled=true)。
/// 已存在的项仅强制 enabled=true,不覆盖其他字段。
pub fn import_from_claude(config: &mut MultiAppConfig) -> Result<usize, AppError> {
let text_opt = crate::claude_mcp::read_mcp_json()?;
let Some(text) = text_opt else { return Ok(0) };
let mut changed = normalize_servers_for(config, &AppType::Claude);
let v: Value = serde_json::from_str(&text)
.map_err(|e| AppError::McpValidation(format!("解析 ~/.claude.json 失败: {}", e)))?;
let Some(map) = v.get("mcpServers").and_then(|x| x.as_object()) else {
return Ok(changed);
};
for (id, spec) in map.iter() {
// 校验目标 spec
validate_server_spec(spec)?;
let entry = config
.mcp_for_mut(&AppType::Claude)
.servers
.entry(id.clone());
use std::collections::hash_map::Entry;
match entry {
Entry::Vacant(vac) => {
let mut obj = serde_json::Map::new();
obj.insert(String::from("id"), json!(id));
obj.insert(String::from("name"), json!(id));
obj.insert(String::from("server"), spec.clone());
obj.insert(String::from("enabled"), json!(true));
vac.insert(Value::Object(obj));
changed += 1;
}
Entry::Occupied(mut occ) => {
let value = occ.get_mut();
let Some(existing) = value.as_object_mut() else {
log::warn!("MCP 条目 '{}' 不是 JSON 对象,覆盖为导入数据", id);
let mut obj = serde_json::Map::new();
obj.insert(String::from("id"), json!(id));
obj.insert(String::from("name"), json!(id));
obj.insert(String::from("server"), spec.clone());
obj.insert(String::from("enabled"), json!(true));
occ.insert(Value::Object(obj));
changed += 1;
continue;
};
let mut modified = false;
let prev_enabled = existing
.get("enabled")
.and_then(|b| b.as_bool())
.unwrap_or(false);
if !prev_enabled {
existing.insert(String::from("enabled"), json!(true));
modified = true;
}
if existing.get("server").is_none() {
log::warn!("MCP 条目 '{}' 缺少 server 字段,覆盖为导入数据", id);
existing.insert(String::from("server"), spec.clone());
modified = true;
}
if existing.get("id").is_none() {
log::warn!("MCP 条目 '{}' 缺少 id 字段,自动填充", id);
existing.insert(String::from("id"), json!(id));
modified = true;
}
if modified {
changed += 1;
}
}
}
}
Ok(changed)
}
/// 从 ~/.codex/config.toml 导入 MCP 到 config.jsonCodex 作用域),并将导入项设为 enabled=true。
/// 支持两种 schema[mcp.servers.<id>] 与 [mcp_servers.<id>]。
/// 已存在的项仅强制 enabled=true,不覆盖其他字段。
pub fn import_from_codex(config: &mut MultiAppConfig) -> Result<usize, AppError> {
let text = crate::codex_config::read_and_validate_codex_config_text()?;
if text.trim().is_empty() {
return Ok(0);
}
let mut changed_total = normalize_servers_for(config, &AppType::Codex);
let root: toml::Table = toml::from_str(&text)
.map_err(|e| AppError::McpValidation(format!("解析 ~/.codex/config.toml 失败: {}", e)))?;
// helper:处理一组 servers 表
let mut import_servers_tbl = |servers_tbl: &toml::value::Table| {
let mut changed = 0usize;
for (id, entry_val) in servers_tbl.iter() {
let Some(entry_tbl) = entry_val.as_table() else {
continue;
};
// type 缺省为 stdio
let typ = entry_tbl
.get("type")
.and_then(|v| v.as_str())
.unwrap_or("stdio");
// 构建 JSON 规范
let mut spec = serde_json::Map::new();
spec.insert("type".into(), json!(typ));
match typ {
"stdio" => {
if let Some(cmd) = entry_tbl.get("command").and_then(|v| v.as_str()) {
spec.insert("command".into(), json!(cmd));
}
if let Some(args) = entry_tbl.get("args").and_then(|v| v.as_array()) {
let arr = args
.iter()
.filter_map(|x| x.as_str())
.map(|s| json!(s))
.collect::<Vec<_>>();
if !arr.is_empty() {
spec.insert("args".into(), serde_json::Value::Array(arr));
}
}
if let Some(cwd) = entry_tbl.get("cwd").and_then(|v| v.as_str()) {
if !cwd.trim().is_empty() {
spec.insert("cwd".into(), json!(cwd));
}
}
if let Some(env_tbl) = entry_tbl.get("env").and_then(|v| v.as_table()) {
let mut env_json = serde_json::Map::new();
for (k, v) in env_tbl.iter() {
if let Some(sv) = v.as_str() {
env_json.insert(k.clone(), json!(sv));
}
}
if !env_json.is_empty() {
spec.insert("env".into(), serde_json::Value::Object(env_json));
}
}
}
"http" => {
if let Some(url) = entry_tbl.get("url").and_then(|v| v.as_str()) {
spec.insert("url".into(), json!(url));
}
if let Some(headers_tbl) = entry_tbl.get("headers").and_then(|v| v.as_table()) {
let mut headers_json = serde_json::Map::new();
for (k, v) in headers_tbl.iter() {
if let Some(sv) = v.as_str() {
headers_json.insert(k.clone(), json!(sv));
}
}
if !headers_json.is_empty() {
spec.insert("headers".into(), serde_json::Value::Object(headers_json));
}
}
}
_ => {}
}
let spec_v = serde_json::Value::Object(spec);
// 校验
if let Err(e) = validate_server_spec(&spec_v) {
log::warn!("跳过无效 Codex MCP 项 '{}': {}", id, e);
continue;
}
// 合并:仅强制 enabled=true
use std::collections::hash_map::Entry;
let entry = config
.mcp_for_mut(&AppType::Codex)
.servers
.entry(id.clone());
match entry {
Entry::Vacant(vac) => {
let mut obj = serde_json::Map::new();
obj.insert(String::from("id"), json!(id));
obj.insert(String::from("name"), json!(id));
obj.insert(String::from("server"), spec_v.clone());
obj.insert(String::from("enabled"), json!(true));
vac.insert(serde_json::Value::Object(obj));
changed += 1;
}
Entry::Occupied(mut occ) => {
let value = occ.get_mut();
let Some(existing) = value.as_object_mut() else {
log::warn!("MCP 条目 '{}' 不是 JSON 对象,覆盖为导入数据", id);
let mut obj = serde_json::Map::new();
obj.insert(String::from("id"), json!(id));
obj.insert(String::from("name"), json!(id));
obj.insert(String::from("server"), spec_v.clone());
obj.insert(String::from("enabled"), json!(true));
occ.insert(serde_json::Value::Object(obj));
changed += 1;
continue;
};
let mut modified = false;
let prev = existing
.get("enabled")
.and_then(|b| b.as_bool())
.unwrap_or(false);
if !prev {
existing.insert(String::from("enabled"), json!(true));
modified = true;
}
if existing.get("server").is_none() {
log::warn!("MCP 条目 '{}' 缺少 server 字段,覆盖为导入数据", id);
existing.insert(String::from("server"), spec_v.clone());
modified = true;
}
if existing.get("id").is_none() {
log::warn!("MCP 条目 '{}' 缺少 id 字段,自动填充", id);
existing.insert(String::from("id"), json!(id));
modified = true;
}
if modified {
changed += 1;
}
}
}
}
changed
};
// 1) 处理 mcp.servers
if let Some(mcp_val) = root.get("mcp") {
if let Some(mcp_tbl) = mcp_val.as_table() {
if let Some(servers_val) = mcp_tbl.get("servers") {
if let Some(servers_tbl) = servers_val.as_table() {
changed_total += import_servers_tbl(servers_tbl);
}
}
}
}
// 2) 处理 mcp_servers
if let Some(servers_val) = root.get("mcp_servers") {
if let Some(servers_tbl) = servers_val.as_table() {
changed_total += import_servers_tbl(servers_tbl);
}
}
Ok(changed_total)
}
/// 将 config.json 中 Codex 的 enabled==true 项以 TOML 形式写入 ~/.codex/config.toml 的 [mcp.servers]
/// 策略:
/// - 读取现有 config.toml;若语法无效则报错,不尝试覆盖
/// - 仅更新 `mcp.servers` 或 `mcp_servers` 子表,保留 `mcp` 其它键
/// - 仅写入启用项;无启用项时清理对应子表
pub fn sync_enabled_to_codex(config: &MultiAppConfig) -> Result<(), AppError> {
use toml_edit::{DocumentMut, Item, Table};
// 1) 收集启用项(Codex 维度)
let enabled = collect_enabled_servers(&config.mcp.codex);
// 2) 读取现有 config.toml 文本;保持无效 TOML 的错误返回(不覆盖文件)
let base_text = crate::codex_config::read_and_validate_codex_config_text()?;
// 3) 使用 toml_edit 解析(允许空文件)
let mut doc: DocumentMut = if base_text.trim().is_empty() {
DocumentMut::default()
} else {
base_text
.parse::<DocumentMut>()
.map_err(|e| AppError::McpValidation(format!("解析 config.toml 失败: {}", e)))?
};
enum Target {
McpServers, // 顶层 mcp_servers
McpDotServers, // mcp.servers
}
// 4) 选择目标风格:优先沿用既有子表;其次在 mcp 表下新建;最后退回顶层 mcp_servers
let has_mcp_dot_servers = doc
.get("mcp")
.and_then(|m| m.get("servers"))
.and_then(|s| s.as_table_like())
.is_some();
let has_mcp_servers = doc
.get("mcp_servers")
.and_then(|s| s.as_table_like())
.is_some();
let mcp_is_table = doc.get("mcp").and_then(|m| m.as_table_like()).is_some();
let target = if has_mcp_dot_servers {
Target::McpDotServers
} else if has_mcp_servers {
Target::McpServers
} else if mcp_is_table {
Target::McpDotServers
} else {
Target::McpServers
};
// 构造目标 servers 表(稳定的键顺序)
let build_servers_table = || -> Table {
let mut servers = Table::new();
let mut ids: Vec<_> = enabled.keys().cloned().collect();
ids.sort();
for id in ids {
let spec = enabled.get(&id).expect("spec must exist");
let mut t = Table::new();
let typ = spec.get("type").and_then(|v| v.as_str()).unwrap_or("stdio");
t["type"] = toml_edit::value(typ);
match typ {
"stdio" => {
let cmd = spec.get("command").and_then(|v| v.as_str()).unwrap_or("");
t["command"] = toml_edit::value(cmd);
if let Some(args) = spec.get("args").and_then(|v| v.as_array()) {
let mut arr_v = toml_edit::Array::default();
for a in args.iter().filter_map(|x| x.as_str()) {
arr_v.push(a);
}
if !arr_v.is_empty() {
t["args"] = toml_edit::Item::Value(toml_edit::Value::Array(arr_v));
}
}
if let Some(cwd) = spec.get("cwd").and_then(|v| v.as_str()) {
if !cwd.trim().is_empty() {
t["cwd"] = toml_edit::value(cwd);
}
}
if let Some(env) = spec.get("env").and_then(|v| v.as_object()) {
let mut env_tbl = Table::new();
for (k, v) in env.iter() {
if let Some(s) = v.as_str() {
env_tbl[&k[..]] = toml_edit::value(s);
}
}
if !env_tbl.is_empty() {
t["env"] = Item::Table(env_tbl);
}
}
}
"http" => {
let url = spec.get("url").and_then(|v| v.as_str()).unwrap_or("");
t["url"] = toml_edit::value(url);
if let Some(headers) = spec.get("headers").and_then(|v| v.as_object()) {
let mut h_tbl = Table::new();
for (k, v) in headers.iter() {
if let Some(s) = v.as_str() {
h_tbl[&k[..]] = toml_edit::value(s);
}
}
if !h_tbl.is_empty() {
t["headers"] = Item::Table(h_tbl);
}
}
}
_ => {}
}
servers[&id[..]] = Item::Table(t);
}
servers
};
// 5) 应用更新:仅就地更新目标子表;避免改动其它键/注释/空白
if enabled.is_empty() {
// 无启用项:移除两种 servers 表(如果存在),但保留 mcp 其它字段
if let Some(mcp_item) = doc.get_mut("mcp") {
if let Some(tbl) = mcp_item.as_table_like_mut() {
tbl.remove("servers");
}
}
doc.as_table_mut().remove("mcp_servers");
} else {
let servers_tbl = build_servers_table();
match target {
Target::McpDotServers => {
// 确保 mcp 为表
if doc.get("mcp").and_then(|m| m.as_table_like()).is_none() {
doc["mcp"] = Item::Table(Table::new());
}
doc["mcp"]["servers"] = Item::Table(servers_tbl);
// 去重:若存在顶层 mcp_servers,则移除以避免重复定义
doc.as_table_mut().remove("mcp_servers");
}
Target::McpServers => {
doc["mcp_servers"] = Item::Table(servers_tbl);
// 去重:若存在 mcp.servers,则移除该子表,保留 mcp 其它键
if let Some(mcp_item) = doc.get_mut("mcp") {
if let Some(tbl) = mcp_item.as_table_like_mut() {
tbl.remove("servers");
}
}
}
}
}
// 6) 写回(仅改 TOML,不触碰 auth.json);toml_edit 会尽量保留未改区域的注释/空白/顺序
let new_text = doc.to_string();
let path = crate::codex_config::get_codex_config_path();
crate::config::write_text_file(&path, &new_text)?;
Ok(())
}
+131
View File
@@ -0,0 +1,131 @@
//! Claude MCP 同步和导入模块
use serde_json::Value;
use std::collections::HashMap;
use crate::app_config::{McpApps, McpConfig, McpServer, MultiAppConfig};
use crate::error::AppError;
use super::validation::{extract_server_spec, validate_server_spec};
/// 返回已启用的 MCP 服务器(过滤 enabled==true
fn collect_enabled_servers(cfg: &McpConfig) -> HashMap<String, Value> {
let mut out = HashMap::new();
for (id, entry) in cfg.servers.iter() {
let enabled = entry
.get("enabled")
.and_then(|v| v.as_bool())
.unwrap_or(false);
if !enabled {
continue;
}
match extract_server_spec(entry) {
Ok(spec) => {
out.insert(id.clone(), spec);
}
Err(err) => {
log::warn!("跳过无效的 MCP 条目 '{id}': {err}");
}
}
}
out
}
/// 将 config.json 中 enabled==true 的项投影写入 ~/.claude.json
pub fn sync_enabled_to_claude(config: &MultiAppConfig) -> Result<(), AppError> {
let enabled = collect_enabled_servers(&config.mcp.claude);
crate::claude_mcp::set_mcp_servers_map(&enabled)
}
/// 从 ~/.claude.json 导入 mcpServers 到统一结构(v3.7.0+
/// 已存在的服务器将启用 Claude 应用,不覆盖其他字段和应用状态
pub fn import_from_claude(config: &mut MultiAppConfig) -> Result<usize, AppError> {
let text_opt = crate::claude_mcp::read_mcp_json()?;
let Some(text) = text_opt else { return Ok(0) };
let v: Value = serde_json::from_str(&text)
.map_err(|e| AppError::McpValidation(format!("解析 ~/.claude.json 失败: {e}")))?;
let Some(map) = v.get("mcpServers").and_then(|x| x.as_object()) else {
return Ok(0);
};
// 确保新结构存在
let servers = config.mcp.servers.get_or_insert_with(HashMap::new);
let mut changed = 0;
let mut errors = Vec::new();
for (id, spec) in map.iter() {
// 校验:单项失败不中止,收集错误继续处理
if let Err(e) = validate_server_spec(spec) {
log::warn!("跳过无效 MCP 服务器 '{id}': {e}");
errors.push(format!("{id}: {e}"));
continue;
}
if let Some(existing) = servers.get_mut(id) {
// 已存在:仅启用 Claude 应用
if !existing.apps.claude {
existing.apps.claude = true;
changed += 1;
log::info!("MCP 服务器 '{id}' 已启用 Claude 应用");
}
} else {
// 新建服务器:默认仅启用 Claude
servers.insert(
id.clone(),
McpServer {
id: id.clone(),
name: id.clone(),
server: spec.clone(),
apps: McpApps {
claude: true,
codex: false,
gemini: false,
},
description: None,
homepage: None,
docs: None,
tags: Vec::new(),
},
);
changed += 1;
log::info!("导入新 MCP 服务器 '{id}'");
}
}
if !errors.is_empty() {
log::warn!("导入完成,但有 {} 项失败: {:?}", errors.len(), errors);
}
Ok(changed)
}
/// 将单个 MCP 服务器同步到 Claude live 配置
pub fn sync_single_server_to_claude(
_config: &MultiAppConfig,
id: &str,
server_spec: &Value,
) -> Result<(), AppError> {
// 读取现有的 MCP 配置
let current = crate::claude_mcp::read_mcp_servers_map()?;
// 创建新的 HashMap,包含现有的所有服务器 + 当前要同步的服务器
let mut updated = current;
updated.insert(id.to_string(), server_spec.clone());
// 写回
crate::claude_mcp::set_mcp_servers_map(&updated)
}
/// 从 Claude live 配置中移除单个 MCP 服务器
pub fn remove_server_from_claude(id: &str) -> Result<(), AppError> {
// 读取现有的 MCP 配置
let mut current = crate::claude_mcp::read_mcp_servers_map()?;
// 移除指定服务器
current.remove(id);
// 写回
crate::claude_mcp::set_mcp_servers_map(&current)
}
+651
View File
@@ -0,0 +1,651 @@
//! Codex MCP 同步和导入模块
//!
//! 包含 Codex 的 MCP 配置管理:
//! - 从 ~/.codex/config.toml 导入
//! - 同步到 ~/.codex/config.toml
//! - JSON 到 TOML 的转换逻辑
use serde_json::{json, Value};
use std::collections::HashMap;
use crate::app_config::{McpApps, McpConfig, McpServer, MultiAppConfig};
use crate::error::AppError;
use super::validation::{extract_server_spec, validate_server_spec};
/// 返回已启用的 MCP 服务器(过滤 enabled==true
fn collect_enabled_servers(cfg: &McpConfig) -> HashMap<String, Value> {
let mut out = HashMap::new();
for (id, entry) in cfg.servers.iter() {
let enabled = entry
.get("enabled")
.and_then(|v| v.as_bool())
.unwrap_or(false);
if !enabled {
continue;
}
match extract_server_spec(entry) {
Ok(spec) => {
out.insert(id.clone(), spec);
}
Err(err) => {
log::warn!("跳过无效的 MCP 条目 '{id}': {err}");
}
}
}
out
}
/// 从 ~/.codex/config.toml 导入 MCP 到统一结构(v3.7.0+
///
/// 格式支持:
/// - 正确格式:[mcp_servers.*]Codex 官方标准)
/// - 错误格式:[mcp.servers.*](容错读取,用于迁移错误写入的配置)
///
/// 已存在的服务器将启用 Codex 应用,不覆盖其他字段和应用状态
pub fn import_from_codex(config: &mut MultiAppConfig) -> Result<usize, AppError> {
let text = crate::codex_config::read_and_validate_codex_config_text()?;
if text.trim().is_empty() {
return Ok(0);
}
let root: toml::Table = toml::from_str(&text)
.map_err(|e| AppError::McpValidation(format!("解析 ~/.codex/config.toml 失败: {e}")))?;
// 确保新结构存在
let servers = config.mcp.servers.get_or_insert_with(HashMap::new);
let mut changed_total = 0usize;
// helper:处理一组 servers 表
let mut import_servers_tbl = |servers_tbl: &toml::value::Table| {
let mut changed = 0usize;
for (id, entry_val) in servers_tbl.iter() {
let Some(entry_tbl) = entry_val.as_table() else {
continue;
};
// type 缺省为 stdio
let typ = entry_tbl
.get("type")
.and_then(|v| v.as_str())
.unwrap_or("stdio");
// 构建 JSON 规范
let mut spec = serde_json::Map::new();
spec.insert("type".into(), json!(typ));
// 核心字段(需要手动处理的字段)
let core_fields = match typ {
"stdio" => vec!["type", "command", "args", "env", "cwd"],
"http" | "sse" => vec!["type", "url", "http_headers"],
_ => vec!["type"],
};
// 1. 处理核心字段(强类型)
match typ {
"stdio" => {
if let Some(cmd) = entry_tbl.get("command").and_then(|v| v.as_str()) {
spec.insert("command".into(), json!(cmd));
}
if let Some(args) = entry_tbl.get("args").and_then(|v| v.as_array()) {
let arr = args
.iter()
.filter_map(|x| x.as_str())
.map(|s| json!(s))
.collect::<Vec<_>>();
if !arr.is_empty() {
spec.insert("args".into(), serde_json::Value::Array(arr));
}
}
if let Some(cwd) = entry_tbl.get("cwd").and_then(|v| v.as_str()) {
if !cwd.trim().is_empty() {
spec.insert("cwd".into(), json!(cwd));
}
}
if let Some(env_tbl) = entry_tbl.get("env").and_then(|v| v.as_table()) {
let mut env_json = serde_json::Map::new();
for (k, v) in env_tbl.iter() {
if let Some(sv) = v.as_str() {
env_json.insert(k.clone(), json!(sv));
}
}
if !env_json.is_empty() {
spec.insert("env".into(), serde_json::Value::Object(env_json));
}
}
}
"http" | "sse" => {
if let Some(url) = entry_tbl.get("url").and_then(|v| v.as_str()) {
spec.insert("url".into(), json!(url));
}
// Read from http_headers (correct Codex format) or headers (legacy) with priority to http_headers
let headers_tbl = entry_tbl
.get("http_headers")
.and_then(|v| v.as_table())
.or_else(|| entry_tbl.get("headers").and_then(|v| v.as_table()));
if let Some(headers_tbl) = headers_tbl {
let mut headers_json = serde_json::Map::new();
for (k, v) in headers_tbl.iter() {
if let Some(sv) = v.as_str() {
headers_json.insert(k.clone(), json!(sv));
}
}
if !headers_json.is_empty() {
spec.insert("headers".into(), serde_json::Value::Object(headers_json));
}
}
}
_ => {
log::warn!("跳过未知类型 '{typ}' 的 Codex MCP 项 '{id}'");
return changed;
}
}
// 2. 处理扩展字段和其他未知字段(通用 TOML → JSON 转换)
for (key, toml_val) in entry_tbl.iter() {
// 跳过已处理的核心字段
if core_fields.contains(&key.as_str()) {
continue;
}
// 通用 TOML 值到 JSON 值转换
let json_val = match toml_val {
toml::Value::String(s) => Some(json!(s)),
toml::Value::Integer(i) => Some(json!(i)),
toml::Value::Float(f) => Some(json!(f)),
toml::Value::Boolean(b) => Some(json!(b)),
toml::Value::Array(arr) => {
// 只支持简单类型数组
let json_arr: Vec<serde_json::Value> = arr
.iter()
.filter_map(|item| match item {
toml::Value::String(s) => Some(json!(s)),
toml::Value::Integer(i) => Some(json!(i)),
toml::Value::Float(f) => Some(json!(f)),
toml::Value::Boolean(b) => Some(json!(b)),
_ => None,
})
.collect();
if !json_arr.is_empty() {
Some(serde_json::Value::Array(json_arr))
} else {
log::debug!("跳过复杂数组字段 '{key}' (TOML → JSON)");
None
}
}
toml::Value::Table(tbl) => {
// 浅层表转为 JSON 对象(仅支持字符串值)
let mut json_obj = serde_json::Map::new();
for (k, v) in tbl.iter() {
if let Some(s) = v.as_str() {
json_obj.insert(k.clone(), json!(s));
}
}
if !json_obj.is_empty() {
Some(serde_json::Value::Object(json_obj))
} else {
log::debug!("跳过复杂对象字段 '{key}' (TOML → JSON)");
None
}
}
toml::Value::Datetime(_) => {
log::debug!("跳过日期时间字段 '{key}' (TOML → JSON)");
None
}
};
if let Some(val) = json_val {
spec.insert(key.clone(), val);
log::debug!("导入扩展字段 '{key}' = {toml_val:?}");
}
}
let spec_v = serde_json::Value::Object(spec);
// 校验:单项失败继续处理
if let Err(e) = validate_server_spec(&spec_v) {
log::warn!("跳过无效 Codex MCP 项 '{id}': {e}");
continue;
}
if let Some(existing) = servers.get_mut(id) {
// 已存在:仅启用 Codex 应用
if !existing.apps.codex {
existing.apps.codex = true;
changed += 1;
log::info!("MCP 服务器 '{id}' 已启用 Codex 应用");
}
} else {
// 新建服务器:默认仅启用 Codex
servers.insert(
id.clone(),
McpServer {
id: id.clone(),
name: id.clone(),
server: spec_v,
apps: McpApps {
claude: false,
codex: true,
gemini: false,
},
description: None,
homepage: None,
docs: None,
tags: Vec::new(),
},
);
changed += 1;
log::info!("导入新 MCP 服务器 '{id}'");
}
}
changed
};
// 1) 处理 mcp.servers
if let Some(mcp_val) = root.get("mcp") {
if let Some(mcp_tbl) = mcp_val.as_table() {
if let Some(servers_val) = mcp_tbl.get("servers") {
if let Some(servers_tbl) = servers_val.as_table() {
changed_total += import_servers_tbl(servers_tbl);
}
}
}
}
// 2) 处理 mcp_servers
if let Some(servers_val) = root.get("mcp_servers") {
if let Some(servers_tbl) = servers_val.as_table() {
changed_total += import_servers_tbl(servers_tbl);
}
}
Ok(changed_total)
}
/// 将 config.json 中 Codex 的 enabled==true 项以 TOML 形式写入 ~/.codex/config.toml
///
/// 格式策略:
/// - 唯一正确格式:[mcp_servers] 顶层表(Codex 官方标准)
/// - 自动清理错误格式:[mcp.servers](如果存在)
/// - 读取现有 config.toml;若语法无效则报错,不尝试覆盖
/// - 仅更新 `mcp_servers` 表,保留其它键
/// - 仅写入启用项;无启用项时清理 mcp_servers 表
pub fn sync_enabled_to_codex(config: &MultiAppConfig) -> Result<(), AppError> {
use toml_edit::{Item, Table};
// 1) 收集启用项(Codex 维度)
let enabled = collect_enabled_servers(&config.mcp.codex);
// 2) 读取现有 config.toml 文本;保持无效 TOML 的错误返回(不覆盖文件)
let base_text = crate::codex_config::read_and_validate_codex_config_text()?;
// 3) 使用 toml_edit 解析(允许空文件)
let mut doc = if base_text.trim().is_empty() {
toml_edit::DocumentMut::default()
} else {
base_text
.parse::<toml_edit::DocumentMut>()
.map_err(|e| AppError::McpValidation(format!("解析 config.toml 失败: {e}")))?
};
// 4) 清理可能存在的错误格式 [mcp.servers]
if let Some(mcp_item) = doc.get_mut("mcp") {
if let Some(tbl) = mcp_item.as_table_like_mut() {
if tbl.contains_key("servers") {
log::warn!("检测到错误的 MCP 格式 [mcp.servers],正在清理并迁移到 [mcp_servers]");
tbl.remove("servers");
}
}
}
// 5) 构造目标 servers 表(稳定的键顺序)
if enabled.is_empty() {
// 无启用项:移除 mcp_servers 表
doc.as_table_mut().remove("mcp_servers");
} else {
// 构建 servers 表
let mut servers_tbl = Table::new();
let mut ids: Vec<_> = enabled.keys().cloned().collect();
ids.sort();
for id in ids {
let spec = enabled.get(&id).expect("spec must exist");
// 复用通用转换函数(已包含扩展字段支持)
match json_server_to_toml_table(spec) {
Ok(table) => {
servers_tbl[&id[..]] = Item::Table(table);
}
Err(err) => {
log::error!("跳过无效的 MCP 服务器 '{id}': {err}");
}
}
}
// 使用唯一正确的格式:[mcp_servers]
doc["mcp_servers"] = Item::Table(servers_tbl);
}
// 6) 写回(仅改 TOML,不触碰 auth.json);toml_edit 会尽量保留未改区域的注释/空白/顺序
let new_text = doc.to_string();
let path = crate::codex_config::get_codex_config_path();
crate::config::write_text_file(&path, &new_text)?;
Ok(())
}
/// 将单个 MCP 服务器同步到 Codex live 配置
/// 始终使用 Codex 官方格式 [mcp_servers],并清理可能存在的错误格式 [mcp.servers]
pub fn sync_single_server_to_codex(
_config: &MultiAppConfig,
id: &str,
server_spec: &Value,
) -> Result<(), AppError> {
use toml_edit::Item;
// 读取现有的 config.toml
let config_path = crate::codex_config::get_codex_config_path();
let mut doc = if config_path.exists() {
let content =
std::fs::read_to_string(&config_path).map_err(|e| AppError::io(&config_path, e))?;
content
.parse::<toml_edit::DocumentMut>()
.map_err(|e| AppError::McpValidation(format!("解析 Codex config.toml 失败: {e}")))?
} else {
toml_edit::DocumentMut::new()
};
// 清理可能存在的错误格式 [mcp.servers]
if let Some(mcp_item) = doc.get_mut("mcp") {
if let Some(tbl) = mcp_item.as_table_like_mut() {
if tbl.contains_key("servers") {
log::warn!("检测到错误的 MCP 格式 [mcp.servers],正在清理并迁移到 [mcp_servers]");
tbl.remove("servers");
}
}
}
// 确保 [mcp_servers] 表存在
if !doc.contains_key("mcp_servers") {
doc["mcp_servers"] = toml_edit::table();
}
// 将 JSON 服务器规范转换为 TOML 表
let toml_table = json_server_to_toml_table(server_spec)?;
// 使用唯一正确的格式:[mcp_servers]
doc["mcp_servers"][id] = Item::Table(toml_table);
// 写回文件
std::fs::write(&config_path, doc.to_string()).map_err(|e| AppError::io(&config_path, e))?;
Ok(())
}
/// 从 Codex live 配置中移除单个 MCP 服务器
/// 从正确的 [mcp_servers] 表中删除,同时清理可能存在于错误位置 [mcp.servers] 的数据
pub fn remove_server_from_codex(id: &str) -> Result<(), AppError> {
let config_path = crate::codex_config::get_codex_config_path();
if !config_path.exists() {
return Ok(()); // 文件不存在,无需删除
}
let content =
std::fs::read_to_string(&config_path).map_err(|e| AppError::io(&config_path, e))?;
let mut doc = content
.parse::<toml_edit::DocumentMut>()
.map_err(|e| AppError::McpValidation(format!("解析 Codex config.toml 失败: {e}")))?;
// 从正确的位置删除:[mcp_servers]
if let Some(mcp_servers) = doc.get_mut("mcp_servers").and_then(|s| s.as_table_mut()) {
mcp_servers.remove(id);
}
// 同时清理可能存在于错误位置的数据:[mcp.servers](如果存在)
if let Some(mcp_table) = doc.get_mut("mcp").and_then(|t| t.as_table_mut()) {
if let Some(servers) = mcp_table.get_mut("servers").and_then(|s| s.as_table_mut()) {
if servers.remove(id).is_some() {
log::warn!("从错误的 MCP 格式 [mcp.servers] 中清理了服务器 '{id}'");
}
}
}
// 写回文件
std::fs::write(&config_path, doc.to_string()).map_err(|e| AppError::io(&config_path, e))?;
Ok(())
}
// ============================================================================
// TOML 转换辅助函数
// ============================================================================
/// 通用 JSON 值到 TOML 值转换器(支持简单类型和浅层嵌套)
///
/// 支持的类型转换:
/// - String → TOML String
/// - Number (i64) → TOML Integer
/// - Number (f64) → TOML Float
/// - Boolean → TOML Boolean
/// - Array[简单类型] → TOML Array
/// - Object → TOML Inline Table (仅字符串值)
///
/// 不支持的类型(返回 None):
/// - null
/// - 深度嵌套对象
/// - 混合类型数组
fn json_value_to_toml_item(value: &Value, field_name: &str) -> Option<toml_edit::Item> {
use toml_edit::{Array, InlineTable, Item};
match value {
Value::String(s) => Some(toml_edit::value(s.as_str())),
Value::Number(n) => {
if let Some(i) = n.as_i64() {
Some(toml_edit::value(i))
} else if let Some(f) = n.as_f64() {
Some(toml_edit::value(f))
} else {
log::warn!("跳过字段 '{field_name}': 无法转换的数字类型 {n}");
None
}
}
Value::Bool(b) => Some(toml_edit::value(*b)),
Value::Array(arr) => {
// 只支持简单类型的数组(字符串、数字、布尔)
let mut toml_arr = Array::default();
let mut all_same_type = true;
for item in arr {
match item {
Value::String(s) => toml_arr.push(s.as_str()),
Value::Number(n) if n.is_i64() => {
if let Some(i) = n.as_i64() {
toml_arr.push(i);
} else {
all_same_type = false;
break;
}
}
Value::Number(n) if n.is_f64() => {
if let Some(f) = n.as_f64() {
toml_arr.push(f);
} else {
all_same_type = false;
break;
}
}
Value::Bool(b) => toml_arr.push(*b),
_ => {
all_same_type = false;
break;
}
}
}
if all_same_type && !toml_arr.is_empty() {
Some(Item::Value(toml_edit::Value::Array(toml_arr)))
} else {
log::warn!("跳过字段 '{field_name}': 不支持的数组类型(混合类型或嵌套结构)");
None
}
}
Value::Object(obj) => {
// 只支持浅层对象(所有值都是字符串)→ TOML Inline Table
let mut inline_table = InlineTable::new();
let mut all_strings = true;
for (k, v) in obj {
if let Some(s) = v.as_str() {
// InlineTable 需要 Value 类型,toml_edit::value() 返回 Item,需要提取内部的 Value
inline_table.insert(k, s.into());
} else {
all_strings = false;
break;
}
}
if all_strings && !inline_table.is_empty() {
Some(Item::Value(toml_edit::Value::InlineTable(inline_table)))
} else {
log::warn!("跳过字段 '{field_name}': 对象值包含非字符串类型,建议使用子表语法");
None
}
}
Value::Null => {
log::debug!("跳过字段 '{field_name}': TOML 不支持 null 值");
None
}
}
}
/// Helper: 将 JSON MCP 服务器规范转换为 toml_edit::Table
///
/// 策略:
/// 1. 核心字段(type, command, args, url, headers, env, cwd)使用强类型处理
/// 2. 扩展字段(timeout、retry 等)通过白名单列表自动转换
/// 3. 其他未知字段使用通用转换器尝试转换
fn json_server_to_toml_table(spec: &Value) -> Result<toml_edit::Table, AppError> {
use toml_edit::{Array, Item, Table};
let mut t = Table::new();
let typ = spec.get("type").and_then(|v| v.as_str()).unwrap_or("stdio");
t["type"] = toml_edit::value(typ);
// 定义核心字段(已在下方处理,跳过通用转换)
let core_fields = match typ {
"stdio" => vec!["type", "command", "args", "env", "cwd"],
"http" | "sse" => vec!["type", "url", "http_headers"],
_ => vec!["type"],
};
// 定义扩展字段白名单(Codex 常见可选字段)
let extended_fields = [
// 通用字段
"timeout",
"timeout_ms",
"startup_timeout_ms",
"startup_timeout_sec",
"connection_timeout",
"read_timeout",
"debug",
"log_level",
"disabled",
// stdio 特有
"shell",
"encoding",
"working_dir",
"restart_on_exit",
"max_restart_count",
// http/sse 特有
"retry_count",
"max_retry_attempts",
"retry_delay",
"cache_tools_list",
"verify_ssl",
"insecure",
"proxy",
];
// 1. 处理核心字段(强类型)
match typ {
"stdio" => {
let cmd = spec.get("command").and_then(|v| v.as_str()).unwrap_or("");
t["command"] = toml_edit::value(cmd);
if let Some(args) = spec.get("args").and_then(|v| v.as_array()) {
let mut arr_v = Array::default();
for a in args.iter().filter_map(|x| x.as_str()) {
arr_v.push(a);
}
if !arr_v.is_empty() {
t["args"] = Item::Value(toml_edit::Value::Array(arr_v));
}
}
if let Some(cwd) = spec.get("cwd").and_then(|v| v.as_str()) {
if !cwd.trim().is_empty() {
t["cwd"] = toml_edit::value(cwd);
}
}
if let Some(env) = spec.get("env").and_then(|v| v.as_object()) {
let mut env_tbl = Table::new();
for (k, v) in env.iter() {
if let Some(s) = v.as_str() {
env_tbl[&k[..]] = toml_edit::value(s);
}
}
if !env_tbl.is_empty() {
t["env"] = Item::Table(env_tbl);
}
}
}
"http" | "sse" => {
let url = spec.get("url").and_then(|v| v.as_str()).unwrap_or("");
t["url"] = toml_edit::value(url);
if let Some(headers) = spec.get("headers").and_then(|v| v.as_object()) {
let mut h_tbl = Table::new();
for (k, v) in headers.iter() {
if let Some(s) = v.as_str() {
h_tbl[&k[..]] = toml_edit::value(s);
}
}
if !h_tbl.is_empty() {
t["http_headers"] = Item::Table(h_tbl);
}
}
}
_ => {}
}
// 2. 处理扩展字段和其他未知字段
if let Some(obj) = spec.as_object() {
for (key, value) in obj {
// 跳过已处理的核心字段
if core_fields.contains(&key.as_str()) {
continue;
}
// 尝试使用通用转换器
if let Some(toml_item) = json_value_to_toml_item(value, key) {
t[&key[..]] = toml_item;
// 记录扩展字段的处理
if extended_fields.contains(&key.as_str()) {
log::debug!("已转换扩展字段 '{key}' = {value:?}");
} else {
log::info!("已转换自定义字段 '{key}' = {value:?}");
}
}
}
}
Ok(t)
}
+126
View File
@@ -0,0 +1,126 @@
//! Gemini MCP 同步和导入模块
use serde_json::Value;
use std::collections::HashMap;
use crate::app_config::{McpApps, McpConfig, McpServer, MultiAppConfig};
use crate::error::AppError;
use super::validation::{extract_server_spec, validate_server_spec};
/// 返回已启用的 MCP 服务器(过滤 enabled==true
fn collect_enabled_servers(cfg: &McpConfig) -> HashMap<String, Value> {
let mut out = HashMap::new();
for (id, entry) in cfg.servers.iter() {
let enabled = entry
.get("enabled")
.and_then(|v| v.as_bool())
.unwrap_or(false);
if !enabled {
continue;
}
match extract_server_spec(entry) {
Ok(spec) => {
out.insert(id.clone(), spec);
}
Err(err) => {
log::warn!("跳过无效的 MCP 条目 '{id}': {err}");
}
}
}
out
}
/// 将 config.json 中 Gemini 的 enabled==true 项写入 Gemini MCP 配置
pub fn sync_enabled_to_gemini(config: &MultiAppConfig) -> Result<(), AppError> {
let enabled = collect_enabled_servers(&config.mcp.gemini);
crate::gemini_mcp::set_mcp_servers_map(&enabled)
}
/// 从 Gemini MCP 配置导入到统一结构(v3.7.0+)
/// 已存在的服务器将启用 Gemini 应用,不覆盖其他字段和应用状态
pub fn import_from_gemini(config: &mut MultiAppConfig) -> Result<usize, AppError> {
let map = crate::gemini_mcp::read_mcp_servers_map()?;
if map.is_empty() {
return Ok(0);
}
// 确保新结构存在
let servers = config.mcp.servers.get_or_insert_with(HashMap::new);
let mut changed = 0;
let mut errors = Vec::new();
for (id, spec) in map.iter() {
// 校验:单项失败不中止,收集错误继续处理
if let Err(e) = validate_server_spec(spec) {
log::warn!("跳过无效 MCP 服务器 '{id}': {e}");
errors.push(format!("{id}: {e}"));
continue;
}
if let Some(existing) = servers.get_mut(id) {
// 已存在:仅启用 Gemini 应用
if !existing.apps.gemini {
existing.apps.gemini = true;
changed += 1;
log::info!("MCP 服务器 '{id}' 已启用 Gemini 应用");
}
} else {
// 新建服务器:默认仅启用 Gemini
servers.insert(
id.clone(),
McpServer {
id: id.clone(),
name: id.clone(),
server: spec.clone(),
apps: McpApps {
claude: false,
codex: false,
gemini: true,
},
description: None,
homepage: None,
docs: None,
tags: Vec::new(),
},
);
changed += 1;
log::info!("导入新 MCP 服务器 '{id}'");
}
}
if !errors.is_empty() {
log::warn!("导入完成,但有 {} 项失败: {:?}", errors.len(), errors);
}
Ok(changed)
}
/// 将单个 MCP 服务器同步到 Gemini live 配置
pub fn sync_single_server_to_gemini(
_config: &MultiAppConfig,
id: &str,
server_spec: &Value,
) -> Result<(), AppError> {
// 读取现有的 MCP 配置
let mut current = crate::gemini_mcp::read_mcp_servers_map()?;
// 添加/更新当前服务器
current.insert(id.to_string(), server_spec.clone());
// 写回
crate::gemini_mcp::set_mcp_servers_map(&current)
}
/// 从 Gemini live 配置中移除单个 MCP 服务器
pub fn remove_server_from_gemini(id: &str) -> Result<(), AppError> {
// 读取现有的 MCP 配置
let mut current = crate::gemini_mcp::read_mcp_servers_map()?;
// 移除指定服务器
current.remove(id);
// 写回
crate::gemini_mcp::set_mcp_servers_map(&current)
}
+28
View File
@@ -0,0 +1,28 @@
//! MCP (Model Context Protocol) 服务器管理模块
//!
//! 本模块负责 MCP 服务器配置的验证、同步和导入导出。
//!
//! ## 模块结构
//!
//! - `validation` - 服务器配置验证
//! - `claude` - Claude MCP 同步和导入
//! - `codex` - Codex MCP 同步和导入(含 TOML 转换)
//! - `gemini` - Gemini MCP 同步和导入
mod claude;
mod codex;
mod gemini;
mod validation;
// 重新导出公共 API
pub use claude::{
import_from_claude, remove_server_from_claude, sync_enabled_to_claude,
sync_single_server_to_claude,
};
pub use codex::{
import_from_codex, remove_server_from_codex, sync_enabled_to_codex, sync_single_server_to_codex,
};
pub use gemini::{
import_from_gemini, remove_server_from_gemini, sync_enabled_to_gemini,
sync_single_server_to_gemini,
};
+69
View File
@@ -0,0 +1,69 @@
//! MCP 服务器配置验证模块
use serde_json::Value;
use crate::error::AppError;
/// 基础校验:允许 stdio/http/sse;或省略 type(视为 stdio)。对应必填字段存在
pub fn validate_server_spec(spec: &Value) -> Result<(), AppError> {
if !spec.is_object() {
return Err(AppError::McpValidation(
"MCP 服务器连接定义必须为 JSON 对象".into(),
));
}
let t_opt = spec.get("type").and_then(|x| x.as_str());
// 支持三种:stdio/http/sse;若缺省 type 则按 stdio 处理(与社区常见 .mcp.json 一致)
let is_stdio = t_opt.map(|t| t == "stdio").unwrap_or(true);
let is_http = t_opt.map(|t| t == "http").unwrap_or(false);
let is_sse = t_opt.map(|t| t == "sse").unwrap_or(false);
if !(is_stdio || is_http || is_sse) {
return Err(AppError::McpValidation(
"MCP 服务器 type 必须是 'stdio'、'http' 或 'sse'(或省略表示 stdio".into(),
));
}
if is_stdio {
let cmd = spec.get("command").and_then(|x| x.as_str()).unwrap_or("");
if cmd.trim().is_empty() {
return Err(AppError::McpValidation(
"stdio 类型的 MCP 服务器缺少 command 字段".into(),
));
}
}
if is_http {
let url = spec.get("url").and_then(|x| x.as_str()).unwrap_or("");
if url.trim().is_empty() {
return Err(AppError::McpValidation(
"http 类型的 MCP 服务器缺少 url 字段".into(),
));
}
}
if is_sse {
let url = spec.get("url").and_then(|x| x.as_str()).unwrap_or("");
if url.trim().is_empty() {
return Err(AppError::McpValidation(
"sse 类型的 MCP 服务器缺少 url 字段".into(),
));
}
}
Ok(())
}
/// 从 MCP 条目中提取服务器规范
pub fn extract_server_spec(entry: &Value) -> Result<Value, AppError> {
let obj = entry
.as_object()
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目必须为 JSON 对象".into()))?;
let server = obj
.get("server")
.ok_or_else(|| AppError::McpValidation("MCP 服务器条目缺少 server 字段".into()))?;
if !server.is_object() {
return Err(AppError::McpValidation(
"MCP 服务器 server 字段必须为 JSON 对象".into(),
));
}
Ok(server.clone())
}
+16
View File
@@ -0,0 +1,16 @@
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Prompt {
pub id: String,
pub name: String,
pub content: String,
#[serde(skip_serializing_if = "Option::is_none")]
pub description: Option<String>,
#[serde(default)]
pub enabled: bool,
#[serde(rename = "createdAt", skip_serializing_if = "Option::is_none")]
pub created_at: Option<i64>,
#[serde(rename = "updatedAt", skip_serializing_if = "Option::is_none")]
pub updated_at: Option<i64>,
}
+41
View File
@@ -0,0 +1,41 @@
use std::path::PathBuf;
use crate::app_config::AppType;
use crate::codex_config::get_codex_auth_path;
use crate::config::get_claude_settings_path;
use crate::error::AppError;
use crate::gemini_config::get_gemini_dir;
/// 返回指定应用所使用的提示词文件路径。
pub fn prompt_file_path(app: &AppType) -> Result<PathBuf, AppError> {
let base_dir: PathBuf = match app {
AppType::Claude => get_base_dir_with_fallback(get_claude_settings_path(), ".claude")?,
AppType::Codex => get_base_dir_with_fallback(get_codex_auth_path(), ".codex")?,
AppType::Gemini => get_gemini_dir(),
};
let filename = match app {
AppType::Claude => "CLAUDE.md",
AppType::Codex => "AGENTS.md",
AppType::Gemini => "GEMINI.md",
};
Ok(base_dir.join(filename))
}
fn get_base_dir_with_fallback(
primary_path: PathBuf,
fallback_dir: &str,
) -> Result<PathBuf, AppError> {
primary_path
.parent()
.map(|p| p.to_path_buf())
.or_else(|| dirs::home_dir().map(|h| h.join(fallback_dir)))
.ok_or_else(|| {
AppError::localized(
"home_dir_not_found",
format!("无法确定 {fallback_dir} 配置目录:用户主目录不存在"),
format!("Cannot determine {fallback_dir} config directory: user home not found"),
)
})
}
+49 -4
View File
@@ -1,3 +1,4 @@
use indexmap::IndexMap;
use serde::{Deserialize, Serialize};
use serde_json::Value;
use std::collections::HashMap;
@@ -22,9 +23,23 @@ pub struct Provider {
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "sortIndex")]
pub sort_index: Option<usize>,
/// 备注信息
#[serde(skip_serializing_if = "Option::is_none")]
pub notes: Option<String>,
/// 供应商元数据(不写入 live 配置,仅存于 ~/.cc-switch/config.json
#[serde(skip_serializing_if = "Option::is_none")]
pub meta: Option<ProviderMeta>,
/// 图标名称(如 "openai", "anthropic"
#[serde(skip_serializing_if = "Option::is_none")]
pub icon: Option<String>,
/// 图标颜色(Hex 格式,如 "#00A67E"
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "iconColor")]
pub icon_color: Option<String>,
/// 是否为代理目标(数据库专用字段,不写入配置文件)
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "isProxyTarget")]
pub is_proxy_target: Option<bool>,
}
impl Provider {
@@ -43,7 +58,11 @@ impl Provider {
category: None,
created_at: None,
sort_index: None,
notes: None,
meta: None,
icon: None,
icon_color: None,
is_proxy_target: None,
}
}
}
@@ -51,7 +70,7 @@ impl Provider {
/// 供应商管理器
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct ProviderManager {
pub providers: HashMap<String, Provider>,
pub providers: IndexMap<String, Provider>,
pub current: String,
}
@@ -63,11 +82,19 @@ pub struct UsageScript {
pub code: String,
#[serde(skip_serializing_if = "Option::is_none")]
pub timeout: Option<u64>,
/// 访问令牌(用于需要登录的接口
/// 用量查询专用的 API Key(通用模板使用
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "apiKey")]
pub api_key: Option<String>,
/// 用量查询专用的 Base URL(通用和 NewAPI 模板使用)
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "baseUrl")]
pub base_url: Option<String>,
/// 访问令牌(用于需要登录的接口,NewAPI 模板使用)
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "accessToken")]
pub access_token: Option<String>,
/// 用户ID(用于需要用户标识的接口)
/// 用户ID(用于需要用户标识的接口NewAPI 模板使用
#[serde(skip_serializing_if = "Option::is_none")]
#[serde(rename = "userId")]
pub user_id: Option<String>,
@@ -120,11 +147,29 @@ pub struct ProviderMeta {
/// 用量查询脚本配置
#[serde(skip_serializing_if = "Option::is_none")]
pub usage_script: Option<UsageScript>,
/// 合作伙伴标记(前端使用 isPartner,保持字段名一致)
#[serde(rename = "isPartner", skip_serializing_if = "Option::is_none")]
pub is_partner: Option<bool>,
/// 合作伙伴促销 key,用于识别 PackyCode 等特殊供应商
#[serde(
rename = "partnerPromotionKey",
skip_serializing_if = "Option::is_none"
)]
pub partner_promotion_key: Option<String>,
/// 成本倍数(用于计算实际成本)
#[serde(rename = "costMultiplier", skip_serializing_if = "Option::is_none")]
pub cost_multiplier: Option<String>,
/// 每日消费限额(USD
#[serde(rename = "limitDailyUsd", skip_serializing_if = "Option::is_none")]
pub limit_daily_usd: Option<String>,
/// 每月消费限额(USD
#[serde(rename = "limitMonthlyUsd", skip_serializing_if = "Option::is_none")]
pub limit_monthly_usd: Option<String>,
}
impl ProviderManager {
/// 获取所有供应商
pub fn get_all_providers(&self) -> &HashMap<String, Provider> {
pub fn get_all_providers(&self) -> &IndexMap<String, Provider> {
&self.providers
}
}
+238
View File
@@ -0,0 +1,238 @@
use once_cell::sync::Lazy;
use std::collections::HashMap;
/// 供应商图标信息
#[derive(Debug, Clone)]
#[allow(dead_code)]
pub struct ProviderIcon {
pub name: &'static str,
pub color: &'static str,
}
/// 供应商名称到图标的默认映射
#[allow(dead_code)]
pub static DEFAULT_PROVIDER_ICONS: Lazy<HashMap<&'static str, ProviderIcon>> = Lazy::new(|| {
let mut m = HashMap::new();
// AI 服务商
m.insert(
"openai",
ProviderIcon {
name: "openai",
color: "#00A67E",
},
);
m.insert(
"anthropic",
ProviderIcon {
name: "anthropic",
color: "#D4915D",
},
);
m.insert(
"claude",
ProviderIcon {
name: "claude",
color: "#D4915D",
},
);
m.insert(
"google",
ProviderIcon {
name: "google",
color: "#4285F4",
},
);
m.insert(
"gemini",
ProviderIcon {
name: "gemini",
color: "#4285F4",
},
);
m.insert(
"deepseek",
ProviderIcon {
name: "deepseek",
color: "#1E88E5",
},
);
m.insert(
"kimi",
ProviderIcon {
name: "kimi",
color: "#6366F1",
},
);
m.insert(
"moonshot",
ProviderIcon {
name: "moonshot",
color: "#6366F1",
},
);
m.insert(
"zhipu",
ProviderIcon {
name: "zhipu",
color: "#0F62FE",
},
);
m.insert(
"minimax",
ProviderIcon {
name: "minimax",
color: "#FF6B6B",
},
);
m.insert(
"baidu",
ProviderIcon {
name: "baidu",
color: "#2932E1",
},
);
m.insert(
"alibaba",
ProviderIcon {
name: "alibaba",
color: "#FF6A00",
},
);
m.insert(
"tencent",
ProviderIcon {
name: "tencent",
color: "#00A4FF",
},
);
m.insert(
"meta",
ProviderIcon {
name: "meta",
color: "#0081FB",
},
);
m.insert(
"microsoft",
ProviderIcon {
name: "microsoft",
color: "#00A4EF",
},
);
m.insert(
"cohere",
ProviderIcon {
name: "cohere",
color: "#39594D",
},
);
m.insert(
"perplexity",
ProviderIcon {
name: "perplexity",
color: "#20808D",
},
);
m.insert(
"mistral",
ProviderIcon {
name: "mistral",
color: "#FF7000",
},
);
m.insert(
"huggingface",
ProviderIcon {
name: "huggingface",
color: "#FFD21E",
},
);
// 云平台
m.insert(
"aws",
ProviderIcon {
name: "aws",
color: "#FF9900",
},
);
m.insert(
"azure",
ProviderIcon {
name: "azure",
color: "#0078D4",
},
);
m.insert(
"huawei",
ProviderIcon {
name: "huawei",
color: "#FF0000",
},
);
m.insert(
"cloudflare",
ProviderIcon {
name: "cloudflare",
color: "#F38020",
},
);
m
});
/// 根据供应商名称智能推断图标
#[allow(dead_code)]
pub fn infer_provider_icon(provider_name: &str) -> Option<ProviderIcon> {
let name_lower = provider_name.to_lowercase();
// 精确匹配
if let Some(icon) = DEFAULT_PROVIDER_ICONS.get(name_lower.as_str()) {
return Some(icon.clone());
}
// 模糊匹配(包含关键词)
for (key, icon) in DEFAULT_PROVIDER_ICONS.iter() {
if name_lower.contains(key) {
return Some(icon.clone());
}
}
None
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_exact_match() {
let icon = infer_provider_icon("openai");
assert!(icon.is_some());
let icon = icon.unwrap();
assert_eq!(icon.name, "openai");
assert_eq!(icon.color, "#00A67E");
}
#[test]
fn test_fuzzy_match() {
let icon = infer_provider_icon("OpenAI Official");
assert!(icon.is_some());
let icon = icon.unwrap();
assert_eq!(icon.name, "openai");
}
#[test]
fn test_case_insensitive() {
let icon = infer_provider_icon("ANTHROPIC");
assert!(icon.is_some());
assert_eq!(icon.unwrap().name, "anthropic");
}
#[test]
fn test_no_match() {
let icon = infer_provider_icon("unknown provider");
assert!(icon.is_none());
}
}
+183
View File
@@ -0,0 +1,183 @@
use axum::{
http::StatusCode,
response::{IntoResponse, Response},
Json,
};
use serde_json::json;
use thiserror::Error;
#[derive(Debug, Error)]
pub enum ProxyError {
#[error("服务器已在运行")]
AlreadyRunning,
#[error("服务器未运行")]
NotRunning,
#[error("地址绑定失败: {0}")]
BindFailed(String),
#[error("请求转发失败: {0}")]
ForwardFailed(String),
#[error("无可用的Provider")]
NoAvailableProvider,
#[allow(dead_code)]
#[error("Provider不健康: {0}")]
ProviderUnhealthy(String),
#[error("上游错误 (状态码 {status}): {body:?}")]
UpstreamError { status: u16, body: Option<String> },
#[error("超过最大重试次数")]
MaxRetriesExceeded,
#[error("数据库错误: {0}")]
DatabaseError(String),
#[error("配置错误: {0}")]
ConfigError(String),
#[allow(dead_code)]
#[error("格式转换错误: {0}")]
TransformError(String),
#[allow(dead_code)]
#[error("无效的请求: {0}")]
InvalidRequest(String),
#[error("超时: {0}")]
Timeout(String),
/// 流式响应空闲超时
#[allow(dead_code)]
#[error("流式响应空闲超时: {0}秒无数据")]
StreamIdleTimeout(u64),
/// 认证错误
#[allow(dead_code)]
#[error("认证失败: {0}")]
AuthError(String),
#[allow(dead_code)]
#[error("内部错误: {0}")]
Internal(String),
}
impl IntoResponse for ProxyError {
fn into_response(self) -> Response {
let (status, body) = match &self {
ProxyError::UpstreamError {
status: upstream_status,
body: upstream_body,
} => {
let http_status =
StatusCode::from_u16(*upstream_status).unwrap_or(StatusCode::BAD_GATEWAY);
// 尝试解析上游响应体为 JSON,如果失败则包装为字符串
let error_body = if let Some(body_str) = upstream_body {
if let Ok(json_body) = serde_json::from_str::<serde_json::Value>(body_str) {
// 上游返回的是 JSON,直接透传
json_body
} else {
// 上游返回的不是 JSON,包装为错误消息
json!({
"error": {
"message": body_str,
"type": "upstream_error",
}
})
}
} else {
json!({
"error": {
"message": format!("Upstream error (status {})", upstream_status),
"type": "upstream_error",
}
})
};
(http_status, error_body)
}
_ => {
let (http_status, message) = match &self {
ProxyError::AlreadyRunning => (StatusCode::CONFLICT, self.to_string()),
ProxyError::NotRunning => (StatusCode::SERVICE_UNAVAILABLE, self.to_string()),
ProxyError::BindFailed(_) => {
(StatusCode::INTERNAL_SERVER_ERROR, self.to_string())
}
ProxyError::ForwardFailed(_) => (StatusCode::BAD_GATEWAY, self.to_string()),
ProxyError::NoAvailableProvider => {
(StatusCode::SERVICE_UNAVAILABLE, self.to_string())
}
ProxyError::ProviderUnhealthy(_) => {
(StatusCode::SERVICE_UNAVAILABLE, self.to_string())
}
ProxyError::MaxRetriesExceeded => {
(StatusCode::SERVICE_UNAVAILABLE, self.to_string())
}
ProxyError::DatabaseError(_) => {
(StatusCode::INTERNAL_SERVER_ERROR, self.to_string())
}
ProxyError::ConfigError(_) => (StatusCode::BAD_REQUEST, self.to_string()),
ProxyError::TransformError(_) => {
(StatusCode::UNPROCESSABLE_ENTITY, self.to_string())
}
ProxyError::InvalidRequest(_) => (StatusCode::BAD_REQUEST, self.to_string()),
ProxyError::Timeout(_) => (StatusCode::GATEWAY_TIMEOUT, self.to_string()),
ProxyError::StreamIdleTimeout(_) => {
(StatusCode::GATEWAY_TIMEOUT, self.to_string())
}
ProxyError::AuthError(_) => (StatusCode::UNAUTHORIZED, self.to_string()),
ProxyError::Internal(_) => {
(StatusCode::INTERNAL_SERVER_ERROR, self.to_string())
}
ProxyError::UpstreamError { .. } => unreachable!(),
};
let error_body = json!({
"error": {
"message": message,
"type": "proxy_error",
}
});
(http_status, error_body)
}
};
(status, Json(body)).into_response()
}
}
/// 错误分类
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum ErrorCategory {
/// 可重试错误(网络问题、5xx)
Retryable, // 网络超时、5xx 错误
/// 不可重试错误(4xx、认证失败)
NonRetryable, // 认证失败、参数错误、4xx 错误
#[allow(dead_code)]
ClientAbort, // 客户端主动中断
}
/// 判断错误是否可重试
#[allow(dead_code)]
pub fn categorize_error(error: &reqwest::Error) -> ErrorCategory {
if error.is_timeout() || error.is_connect() {
return ErrorCategory::Retryable;
}
if let Some(status) = error.status() {
if status.is_server_error() {
ErrorCategory::Retryable
} else if status.is_client_error() {
ErrorCategory::NonRetryable
} else {
ErrorCategory::Retryable
}
} else {
ErrorCategory::Retryable
}
}
+328
View File
@@ -0,0 +1,328 @@
//! 请求转发器
//!
//! 负责将请求转发到上游Provider,支持重试和故障转移
use super::{
error::*,
providers::{get_adapter, ProviderAdapter},
router::ProviderRouter,
types::ProxyStatus,
ProxyError,
};
use crate::{app_config::AppType, database::Database, provider::Provider};
use reqwest::{Client, Response};
use serde_json::Value;
use std::sync::Arc;
use std::time::{Duration, Instant};
use tokio::sync::RwLock;
pub struct RequestForwarder {
client: Client,
router: ProviderRouter,
max_retries: u8,
status: Arc<RwLock<ProxyStatus>>,
}
impl RequestForwarder {
pub fn new(
db: Arc<Database>,
timeout_secs: u64,
max_retries: u8,
status: Arc<RwLock<ProxyStatus>>,
) -> Self {
let mut client_builder = Client::builder();
if timeout_secs > 0 {
client_builder = client_builder.timeout(Duration::from_secs(timeout_secs));
}
let client = client_builder
.build()
.expect("Failed to create HTTP client");
Self {
client,
router: ProviderRouter::new(db),
max_retries,
status,
}
}
/// 转发请求(带重试和故障转移)
pub async fn forward_with_retry(
&self,
app_type: &AppType,
endpoint: &str,
body: Value,
headers: axum::http::HeaderMap,
) -> Result<Response, ProxyError> {
let mut failed_ids = Vec::new();
let mut failover_happened = false;
// 获取适配器
let adapter = get_adapter(app_type);
for attempt in 0..self.max_retries {
// 选择Provider
let provider = self.router.select_provider(app_type, &failed_ids).await?;
log::debug!(
"尝试 {} - 使用Provider: {} ({})",
attempt + 1,
provider.name,
provider.id
);
// 更新状态中的当前Provider信息
{
let mut status = self.status.write().await;
status.current_provider = Some(provider.name.clone());
status.current_provider_id = Some(provider.id.clone());
status.total_requests += 1;
status.last_request_at = Some(chrono::Utc::now().to_rfc3339());
if attempt > 0 {
failover_happened = true;
}
}
let start = Instant::now();
// 转发请求
match self
.forward(&provider, endpoint, &body, &headers, adapter.as_ref())
.await
{
Ok(response) => {
let _latency = start.elapsed().as_millis() as u64;
// 成功:更新健康状态
self.router
.update_health(&provider, app_type, true, None)
.await;
// 更新成功统计
{
let mut status = self.status.write().await;
status.success_requests += 1;
status.last_error = None;
if failover_happened {
status.failover_count += 1;
}
// 重新计算成功率
if status.total_requests > 0 {
status.success_rate = (status.success_requests as f32
/ status.total_requests as f32)
* 100.0;
}
}
return Ok(response);
}
Err(e) => {
let latency = start.elapsed().as_millis() as u64;
// 失败:分类错误
let category = self.categorize_proxy_error(&e);
match category {
ErrorCategory::Retryable => {
// 可重试:更新健康状态,添加到失败列表
self.router
.update_health(&provider, app_type, false, Some(e.to_string()))
.await;
failed_ids.push(provider.id.clone());
// 更新错误信息
{
let mut status = self.status.write().await;
status.last_error =
Some(format!("Provider {} 失败: {}", provider.name, e));
}
log::warn!(
"请求失败(可重试): Provider {} - {} - {}ms",
provider.name,
e,
latency
);
continue;
}
ErrorCategory::NonRetryable | ErrorCategory::ClientAbort => {
// 不可重试:更新失败统计并返回
{
let mut status = self.status.write().await;
status.failed_requests += 1;
status.last_error = Some(e.to_string());
if status.total_requests > 0 {
status.success_rate = (status.success_requests as f32
/ status.total_requests as f32)
* 100.0;
}
}
log::error!("请求失败(不可重试): {e}");
return Err(e);
}
}
}
}
}
// 所有重试都失败
{
let mut status = self.status.write().await;
status.failed_requests += 1;
status.last_error = Some("已达到最大重试次数".to_string());
if status.total_requests > 0 {
status.success_rate =
(status.success_requests as f32 / status.total_requests as f32) * 100.0;
}
}
Err(ProxyError::MaxRetriesExceeded)
}
/// 转发单个请求(使用适配器)
async fn forward(
&self,
provider: &Provider,
endpoint: &str,
body: &Value,
headers: &axum::http::HeaderMap,
adapter: &dyn ProviderAdapter,
) -> Result<Response, ProxyError> {
// 使用适配器提取 base_url
let base_url = adapter.extract_base_url(provider)?;
log::info!("[{}] base_url: {}", adapter.name(), base_url);
// 使用适配器构建 URL
let url = adapter.build_url(&base_url, endpoint);
// 检查是否需要格式转换
let needs_transform = adapter.needs_transform(provider);
// 记录原始请求 JSON
log::info!(
"[{}] ====== 请求开始 ======\n>>> 原始请求 JSON:\n{}",
adapter.name(),
serde_json::to_string_pretty(body).unwrap_or_else(|_| body.to_string())
);
// 转换请求体(如果需要)
let request_body = if needs_transform {
log::info!("[{}] 转换请求格式 (Anthropic → OpenAI)", adapter.name());
let transformed = adapter.transform_request(body.clone(), provider)?;
log::info!(
"[{}] >>> 转换后的请求 JSON:\n{}",
adapter.name(),
serde_json::to_string_pretty(&transformed).unwrap_or_default()
);
transformed
} else {
body.clone()
};
log::info!(
"[{}] 转发请求: {} -> {}",
adapter.name(),
provider.name,
url
);
// 构建请求
let mut request = self.client.post(&url);
// 只透传必要的 Headers(白名单模式)
let allowed_headers = [
"accept",
"user-agent",
"x-request-id",
"x-stainless-arch",
"x-stainless-lang",
"x-stainless-os",
"x-stainless-package-version",
"x-stainless-runtime",
"x-stainless-runtime-version",
];
for (key, value) in headers {
let key_str = key.as_str().to_lowercase();
if allowed_headers.contains(&key_str.as_str()) {
request = request.header(key, value);
}
}
// 确保 Content-Type 是 json
request = request.header("Content-Type", "application/json");
// 使用适配器添加认证头
if let Some(auth) = adapter.extract_auth(provider) {
log::debug!(
"[{}] 使用认证: {:?} (key: {})",
adapter.name(),
auth.strategy,
auth.masked_key()
);
request = adapter.add_auth_headers(request, &auth);
} else {
log::error!(
"[{}] 未找到 API KeyProvider: {}",
adapter.name(),
provider.name
);
}
// 发送请求
log::info!("[{}] 发送请求到: {}", adapter.name(), url);
let response = request.json(&request_body).send().await.map_err(|e| {
log::error!("[{}] 请求失败: {}", adapter.name(), e);
if e.is_timeout() {
ProxyError::Timeout(format!("请求超时: {e}"))
} else if e.is_connect() {
ProxyError::ForwardFailed(format!("连接失败: {e}"))
} else {
ProxyError::ForwardFailed(e.to_string())
}
})?;
// 检查响应状态
let status = response.status();
log::info!("[{}] 响应状态: {}", adapter.name(), status);
if status.is_success() {
Ok(response)
} else {
let status_code = status.as_u16();
let body_text = response.text().await.ok();
log::error!(
"[{}] 上游错误 ({}): {:?}",
adapter.name(),
status_code,
body_text
);
Err(ProxyError::UpstreamError {
status: status_code,
body: body_text,
})
}
}
/// 分类ProxyError
fn categorize_proxy_error(&self, error: &ProxyError) -> ErrorCategory {
match error {
ProxyError::Timeout(_) => ErrorCategory::Retryable,
ProxyError::ForwardFailed(_) => ErrorCategory::Retryable,
ProxyError::UpstreamError { status, .. } => {
if *status >= 500 {
ErrorCategory::Retryable
} else if *status >= 400 && *status < 500 {
ErrorCategory::NonRetryable
} else {
ErrorCategory::Retryable
}
}
ProxyError::ProviderUnhealthy(_) => ErrorCategory::Retryable,
ProxyError::NoAvailableProvider => ErrorCategory::NonRetryable,
_ => ErrorCategory::NonRetryable,
}
}
}
File diff suppressed because it is too large Load Diff

Some files were not shown because too many files have changed in this diff Show More