* feat(db): in-app recovery screen with upgrade button when DB version is too new
When the SQLite user_version is newer than the app supports (SCHEMA_VERSION),
Database::init() fails and previously dead-ended in a native Retry/Exit dialog
(Retry just fails again). The app now boots a dedicated recovery screen instead.
- Detect the recoverable "version too new" case and surface it via init_status
(kind="db_version_too_new" + db_version/supported_version); force-show the main
window and skip normal AppState boot (recovery commands need only AppHandle).
- The recovery screen first checks for an available update:
- update available -> "Upgrade app" runs the updater (download + install +
restart) with a download progress bar.
- no update (already latest) -> warns that the DB is too new even for the latest
build (likely a third-party client), so upgrading cannot help.
- install_update_and_restart now emits `update-download-progress` events; new
`check_app_update_available` command.
- i18n: en / ja / zh / zh-TW.
Verified: pnpm typecheck + format:check + test:unit (351) green; cross-model
review of Rust correctness and recovery-mode safety (no AppState panic).
* fix(db): pre-check too-new DB before schema writes; exit on close in recovery mode
Addresses review feedback on the too-new-DB recovery flow.
- P1: stored_user_version_exceeds_supported() is now checked BEFORE
Database::init(), so create_tables()'s DDL (incl. the unconditional
DROP INDEX/DROP TABLE IF EXISTS failover_queue) never runs against a
database whose user_version we cannot understand. The earlier
post-init-failure recovery branch is removed; the pre-check is the
single authoritative guard, so "don't write to a DB we can't read"
now holds from the very first DB access.
- P2: the window CloseRequested handler now detects recovery mode
(init_error kind = db_version_too_new) and exits the app instead of
honoring minimize_to_tray_on_close. Recovery mode returns before the
tray is created, so hiding the window would leave the app running with
no tray to bring it back; native-close now quits cleanly.
Verified: cargo fmt --check clean; cross-model Rust review of both fixes
(compile-correctness + no normal-startup/close regression).
CI cargo fmt --check failed on transform_codex_chat.rs (the wrapping
introduced upstream did not match rustfmt 1.95.0). Apply cargo fmt so the
chat_legacy_function_call_to_response_item call uses the expected layout.
Query coding-plan / agent-plan usage for Volcengine Ark via the
control-plane OpenAPI (open.volcengineapi.com), which requires
account-level AccessKey signing rather than the inference API key.
- Implement Volcengine Signature V4 (an AWS SigV4 variant: fixed
canonical header order, "HMAC-SHA256" algorithm without the AWS4
prefix, scope ending in "request", service "ark") in
services/coding_plan.rs
- Auto-detect the plan by probing GetAFPUsage (Agent Plan, AFP
five-hour/weekly/monthly quotas) first, falling back to
GetCodingPlanUsage (Coding Plan, session/weekly/monthly percentages);
a single credential covers whichever plan the account holds
- Parse the real response shape: the window label lives in the `Level`
field; guard ResetTimestamp <= 0 (session returns -1)
- Store account-level credentials on UsageScript (accessKeyId /
secretAccessKey), threaded through both the test command and the
TOKEN_PLAN auto-refresh path
- Add an independent AK/SK input block with a detailed hint pointing to
the Volcengine console -> API Access Keys
- Add the `monthly` tier label (frontend TIER_I18N_KEYS + tray
TIER_LABEL_GROUPS + subscription service)
- Sync zh / en / ja / zh-TW locales and the usage-query docs
The vendor renamed its domain, API endpoints, and trademark from CTok to
ETok. Update all user-facing branding accordingly:
- Presets (claude, claude-desktop, codex, gemini, hermes, openclaw,
opencode): ctok.ai -> etok.ai, api.ctok.ai -> api.etok.ai, internal
provider id ctok -> etok, display name CTok.ai -> ETok.ai
- Icons: replace inline placeholder SVG with the new ETok brand PNG,
moved from `icons` to `iconUrls`; key ctok -> etok; banner logo
ctok.png -> etok.png
- README (en/zh/ja/de) partner banner: new logo, links, and copy
- i18n (en/ja/zh/zh-TW): partnerPromotion key ctok -> etok
Keep "ctok" in the Codex history-migration whitelist as a legacy id and
add "etok", so existing users' local session history stays correctly
bucketed after the rename.
* Chat API: skip tool calls with missing function names
Some providers send empty or absent function names in streaming
tool call deltas. Previously these produced invalid output items.
- Don't overwrite accumulated state.name with empty deltas
- Skip tool calls that never received a valid name (instead of
falling back to 'unknown_tool')
- Apply the same defensive guard in finalize_tools and the
non-streaming path
* Address review: defer empty-name skip to finalization
Require both call_id and name before triggering should_add,
instead of skipping eagerly when name is absent in the first
delta. This handles providers that send id before name,
as suggested in the Codex review.
* Guard legacy function_call against empty name
Return Option<Value> from chat_legacy_function_call_to_response_item,
returning None when function_call.name is missing or empty. This covers
the legacy message.function_call path that the original guard missed.
* Remove unreachable unknown_tool fallback
---------
Co-authored-by: Jason <farion1231@gmail.com>
The session-history migration only scanned ~/.codex/state_5.sqlite and the
config.toml `sqlite_home` location. When Codex's SQLite state is relocated via
the CODEX_SQLITE_HOME env var, the state DB was never scanned and its threads
kept their old provider bucket. Add a CODEX_SQLITE_HOME fallback (config
`sqlite_home` still takes precedence) in the shared codex_state_db_paths helper
used by both the third-party and unified-session history migrations, plus serial
env tests.
* feat(usage): support importing model pricing from models.dev
Add an "Import from models.dev" button to the Add Pricing panel that
fetches https://models.dev/api.json, lists priced models sorted by
release date (newest 50 by default, full-text search across ~4800),
and bulk-imports the selected entries through the same
update_model_pricing command as manual entry.
- Normalize imported model IDs to match the backend's
clean_model_id_for_pricing rules (strip vendor prefix, lowercase,
truncate ':' suffix, map '@' to '-', drop the [1m] marker) so the
stored rows actually match cost-attribution lookups
- Dedupe selections that collapse to the same model_id and report
skipped duplicates in the success toast
- Invalidate usage queries on settled (not just success) so partial
import failures still refresh the pricing list
- Keep ESC inside the picker's search input from closing the dialog
and discarding the selection
- Add i18n keys for zh/en/zh-TW/ja and unit tests for the
normalization, price formatting and flattening logic
Fixes#4017
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(usage): match scoped cost backfill against raw model aliases
The scoped backfill selected zero-cost rows via exact SQL string match,
but log columns store raw model strings (route prefixes, :free variants,
date suffixes), so alias rows were skipped until the next full backfill
on startup. Filter rows in Rust with the same model_pricing_candidates
normalization used by the pricing lookup; pricing decision logic is
untouched. Pre-existing gap from schema v11, surfaced by bulk import.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* refactor(usage): restrict models.dev pricing import to a single model
Each update_model_pricing call triggers a backfill pass that loads
every zero-cost usage row before filtering by model, so bulk-importing
N entries scaled as selectedModels x allZeroCostLogs full scans.
Re-importing pricing is rare, so drop the batch path instead of
optimizing it: the picker is now single-select, one import runs exactly
one update_model_pricing call and one backfill pass. This also removes
the normalized-ID dedup logic and the useImportModelPricing hook in
favor of the existing useUpdateModelPricing.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* fix(terminal): detect user default shell instead of hardcoded bash
All macOS/Linux terminal launch functions in misc.rs hardcoded `bash` as
the execution shell, causing zsh users' .zshrc to not load. Now detect
the user's default shell via $SHELL (macOS fallback /bin/zsh, Linux
fallback /bin/bash) and use it consistently across shebang, AppleScript,
spawn args, and interactive shell flags.
- Add get_user_shell(), get_shell_name(), interactive_flags_for_shell()
- Replace 10 hardcoded bash locations with dynamic shell detection
- zsh gets --norcs --no-globalrcs, bash keeps --norc --noprofile
- run_tool_lifecycle_silently unchanged (bash-specific syntax)
- Fixes#1546, Related #2385
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* fix(terminal): use POSIX sh for script execution, user shell for interactive session
Address code review issues:
1. Replace shell-specific `read -n 1 -s`/`read -k 1 -s` with POSIX `read -r _`
(press Enter to close) — ensures portability across bash/zsh/fish/dash
2. Unify AppleScript to use `sh` consistently (not mixing shell/shell_name)
3. Protect fish users: scripts always interpreted by POSIX sh, not fish
(fish doesn't support trap/exec). exec line switches to user's shell.
4. Use build_exec_line to avoid trailing space when interactive_flags is empty
All terminal launchers now execute scripts via `sh`. Script shebangs are
`#!/bin/sh`. The exec line at script end switches to the user's detected
shell with appropriate clean-start flags. Terminal launcher functions no
longer need shell/shell_name parameters — simplified signatures.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* fix: validate user shell for terminal launch
* fix: load user shell config after terminal command
* test: align terminal shell exec expectations
* refactor(terminal): remove dead code, quote Warp path, unify shebangs
- Remove `interactive_flags_for_shell` (always returned "") and
`get_shell_name` (no longer needed after removing flags); simplify
`build_exec_line` to single-arg form
- Quote the Warp launcher script path with `shell_single_quote` for
consistency with the main launcher scripts
- Unify all script shebangs to `#!/usr/bin/env sh` for better
portability (e.g. NixOS where /bin/sh may not exist)
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* fix(terminal): fix linux shell exec helper call
* fix(terminal): run provider command through user shell
* fix(terminal): validate user shell executable
* fix(terminal): avoid interactive provider shell job
* fix(terminal): harden user shell launcher
Load zsh provider commands through the user's login and interactive environment, preserve launch cwd handling, and replace Terminal/iTerm launcher shells so the final user shell is restored cleanly.
---------
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
* fix(codex): restore cached tool call fields
* refactor(codex): merge duplicate enrich loops in chat history
enrich_call_item_from_cache copied the fill-if-empty loop for
reasoning_content/reasoning. The two loops are identical and key
order is irrelevant, so fold both key sets into a single loop.
Pure refactor, no behavior change; codex_chat_history tests pass.
---------
Co-authored-by: Jason <farion1231@gmail.com>
DeepSeek's Anthropic-compatible endpoint rejects requests where
thinking.type=disabled coexists with effort parameters, returning
HTTP 400. This breaks Claude Code 2.1.166+ sub-agents (Workflow/Dynamic
Workflow), which hardcode thinking:disabled.
Rather than overriding thinking:disabled, remove the conflicting effort
parameters (output_config.effort / reasoning_effort) to respect Claude
Code's intent — sub-agents don't need to display reasoning.
Fixes: https://github.com/deepseek-ai/DeepSeek-V3/issues/1397
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Codex ships as an npm platform-dispatch package (JS launcher @openai/codex + per-platform binary optional deps like @openai/codex-darwin-arm64). The upgrade chain ran `<bin>/codex update || <bin>/npm i -g @openai/codex@latest`, which can leave codex throwing "Missing optional dependency @openai/codex-darwin-arm64":
- `codex update` on an npm install is a bare `npm install -g @openai/codex` that prints success and exits 0 even when the platform binary fails to land, short-circuiting the `||` npm fallback.
- The npm fallback is a no-op when version==latest and only targets the main package, so it can never re-land the missing platform binary.
Fixes: (1) remove codex from prefers_official_update (Posix+Windows) so npm-managed codex no longer runs the false-success `codex update`; (2) add a runnable=false gate in installs_anchored_command emitting an uninstall+install self-heal — the only repair that re-lands the platform binary; (3) narrow by source/real to npm-managed sources (nvm/fnm/mise/homebrew, non-brew-formula) so broken brew-formula/volta/bun installs fall back to their own anchored commands instead of being mis-repaired with npm. Reuses the existing enumerate runnable signal; no new FS probing.
- useUsageQuery: retry once + keep-last-good — show the last successful
result for up to 10min when a query fails transiently (network/timeout/
HTTP 5xx), so a single blip no longer flips the card to red. Deterministic
failures (auth, empty key, unknown provider, 4xx) surface immediately and
clear the snapshot so a stale quota can't resurface after credentials change.
- bump native balance/coding-plan/subscription request timeouts 10s -> 15s
for slow cross-border endpoints.
- coding_plan: return explicit errors ("API key is empty" / "Unknown coding
plan provider") instead of a blank failure, mirroring balance.
- add unit tests for keep-last-good and transient/deterministic classification.
Official providers (Claude/Codex/Gemini/Claude Desktop) use OAuth with an intentionally empty base_url and connect via the client's default endpoint, so cc-switch has no reliable reachability target. Probing a guessed endpoint either hits the wrong target or returns a meaningless green light. Hide the connectivity button for category === 'official'; reachability stays available for copilot/codex-oauth/third-party/custom providers, which is where the old real-request probe produced false negatives. Revert the official base_url fallback added earlier — resolve_base_url is back to extract-or-error.
The 1500ms degraded threshold was too strict; normal ~1s probe latencies showed as 'slow'. Restore the original 6000ms scale (default + config panel + per-provider range). Keep the reachability-appropriate 8s timeout / 1 retry.
The provider panel health check sent a real streaming model request, which many third-party providers block (401/403/WAF), causing false negatives while only stable official endpoints passed. Replace it with a lightweight reachability probe: GET the provider base_url and treat any HTTP response (200/4xx/5xx) as reachable; only DNS/connect/TLS/timeout count as failure. Latency is the probe's TTFB.
Backend (services/stream_check.rs): rewrite ~2200 -> ~350 lines, dropping real-request building, format conversion, auth and API-path resolution while keeping per-app base_url extraction. Defaults: 8s timeout, 1 retry, 1500ms degraded threshold.
Failover invariant: the reachability check must never reset the circuit breaker (reachable != usable; a 403 host is reachable but broken for real traffic). Remove the resetCircuitBreaker call from useStreamCheck; failover failure detection stays driven solely by real proxy traffic (forwarder/circuit_breaker untouched). useResetCircuitBreaker is kept dormant for a future manual-recovery entry.
Open the check to all providers: drop the official/copilot/codex-oauth/third-party gating and the 'sends a real request' confirm dialog. For official providers whose base_url is intentionally empty, fall back to the endpoint the client actually uses (Claude -> api.anthropic.com, Codex -> chatgpt.com/backend-api/codex, Gemini -> generativelanguage). Non-official providers with a missing base_url still error to avoid a false green light. Claude Desktop Official is native 1P mode (talks to claude.ai, cc-switch not in the request path, no reliable endpoint) so its button stays hidden.
Slim StreamCheckConfig and per-provider testConfig to timeout/threshold/retries (drop test model + prompt); sync zh/en/ja/zh-TW. Retain the now-unused anthropic_to_openai/anthropic_to_gemini transform utilities and their test suites.
Add kimi-k2.7-code pricing seed (in $0.95 / out $4.00 / cache-read $0.19
per 1M tokens, 256K context) and point all six official Moonshot Kimi
presets (Claude Code, Codex, Claude Desktop, Hermes, OpenCode, OpenClaw)
at the new model. Rename the version-tagged OpenCode/OpenClaw presets to
"Kimi K2.7 Code" and correct the OpenClaw context window to 262144.
The seed is applied via the idempotent INSERT OR IGNORE path that runs on
every startup, so existing users pick up the new pricing without a schema
migration. Kimi For Coding and Nvidia presets are intentionally untouched.
- Wire claude-fable-5 as a fourth tier on both proxy paths, with a
fable -> opus -> default fallback mirroring the official downgrade.
- Whitelist the fable- prefix for the Desktop 1.12603.1+ validator.
- Clarify fallbackModelHint (zh/en/ja/zh-TW): a blank tier on
third-party endpoints forwards the literal model name and 404s.
Refs #3980, #4026, #4049.
- Enable dialog gains a checkbox (default off) to migrate existing
official sessions from the built-in "openai" bucket into the shared
"custom" bucket, with per-generation backups; failed migrations retry
at startup
- Disable dialog offers a precise restore driven by the backup ledger:
only sessions recorded as "openai" in backups are flipped back, and
sessions created while the toggle was on are never touched
- Completion marker and backup generations are bound to the canonical
Codex config dir; migrate/restore serialize on an op lock and the
marker is written conditionally inside the settings write lock
- save_settings rolls back the toggle and fails the save when the live
rewrite fails; migration additionally requires the live config to
actually route to the shared bucket (skips with live_not_unified so
refused injection or proxy takeover can't split history)
- Restore refuses to run while the toggle is (re-)enabled and reports
nothing_to_restore instead of a zero-count success; local migration
markers are now backend-owned in merge_settings_for_save so stale
frontend payloads can't resurrect them
- Settings autosave reverts optimistic form state on failure so a
failed toggle change can't be replayed by an unrelated save
- ConfirmDialog supports an optional checkbox; all four locales updated
Codex buckets resume history by the model_provider id recorded in each
session: official runs (no key, built-in "openai") and cc-switch
third-party runs (shared "custom") are mutually invisible in the resume
picker. Add an opt-in setting that runs official providers under the
shared "custom" id so future official sessions land in the same history
bucket as third-party ones. Forward-only by design: existing sessions
are not migrated.
When enabled, official live config.toml gets model_provider = "custom"
plus a [model_providers.custom] entry that mirrors the built-in openai
provider (requires_openai_auth routes auth to the ChatGPT login in
auth.json, name "OpenAI" keeps is_openai() feature gates, explicit
supports_websockets/wire_api restore built-in defaults). auth.json is
untouched.
Key invariants:
- Injection lives only in the live config: switch-away backfill strips
the exact injected shape, so stored provider configs stay clean and
turning the toggle off fully reverts on the next write.
- Toggle changes apply immediately via a takeover-aware reapply: when
the proxy owns the live config (backup/placeholder present), only the
live backup is updated, mirroring the provider-switch path.
- The takeover backup path runs the same injection so a takeover
release restores a config that still carries the unified routing.
- Injection refuses to activate a foreign [model_providers.custom]
table (e.g. stale entry with a third-party base_url) to avoid routing
ChatGPT OAuth traffic to an unknown backend.
The toggle lives under Settings → Codex App Enhancements; the
description warns that resuming old sessions across providers may fail
because encrypted_content reasoning only decrypts on the backend that
created it (upstream treats cross-provider resume as unsupported).
The provider/model filters only lived inside the request-log table, so
there was no way to see "how much did app X spend on source Y" across
the whole dashboard. Promote them to the top bar next to the app
filter, applying globally to the hero summary, trend chart, request
logs, and both stats tabs.
Backend: the five stats queries (summary, summary-by-app, trends,
provider stats, model stats) accept optional provider_name/model
filters, applied to both the detail and daily-rollup branches (the
rollup PK already carries provider_id/model/pricing_model). Sources
match by exact display name via provider_name_coalesce, so session
placeholder rows like "Claude (Session)" are selectable; models match
by effective pricing model (pricing_model falling back to model), the
same grouping key the model-stats tab uses. Request-log filtering
switches from LIKE to these exact semantics.
Frontend: two truncating dropdowns list only sources/models that have
data in the current range, with the model list cascading from the
selected source. Dynamic option values are prefix-encoded so a source
literally named "all" cannot collide with the sentinel, query keys
fall back to null instead of "all", and the option queries follow the
dashboard refresh interval (otherwise their default 30s polling drags
same-key stats queries along even with refresh disabled). The
request-log filter bar keeps only the log-specific status-code select.
Labels read "sources" rather than "providers" because direct-connect
session buckets sit alongside real providers. i18n updated across
zh/en/ja/zh-TW.
The Desktop gateway's proxy traffic is still recorded under its own
app_type for route-takeover billing audit (the request detail panel
shows the real value), but the dashboard now folds it into `claude`
for display. A standalone "Claude Desktop" bucket only ever showed a
partial number: Desktop chat usage never passes through the proxy and
has no scannable local source, while its Code-tab sessions are just the
embedded Claude Code runtime writing into the shared ~/.claude/projects
tree — so a separate bucket misled users into reading it as Desktop's
full usage.
Backend: new `folded_app_type_sql` helper wraps the app_type column in
every dashboard read path (10 filter sites + the by-app projection) so
`= 'claude'` also matches claude-desktop and GROUP BY merges the two,
without changing bound-param counts. Dedup matching and provider-limit
checks keep exact comparison; get_request_logs folds only the WHERE
filter and keeps the raw app_type in its row projection.
Frontend: drop claude-desktop from the dashboard AppType/KNOWN_APP_TYPES
filter list, the UsageHero theme, and the now-dead appFilter locale key
in all four languages (the managed-app apps.claude-desktop label stays).
Adds test_claude_desktop_folds_into_claude_for_display.
* fix(proxy): preserve Codex OAuth auth token on takeover
* style(proxy): format Codex OAuth takeover fix
* fix(proxy): unconditionally inject AUTH_TOKEN placeholder for codex takeover
The preserve-if-exists condition left #3784 unfixed on three paths:
hot-switch passes the provider's settings (presets carry no
ANTHROPIC_AUTH_TOKEN key), fresh installs never had the key, and live
configs already stripped by older releases stay stripped.
- Fold the bool parameter into the policy enum as
ManagedAccount { keep_auth_token } so every construction site
declares intent
- Decide via !is_github_copilot() within the managed branch so
URL-only codex providers (no provider_type meta) are covered,
matching the predicate family used for policy selection
- Inject the placeholder unconditionally instead of only when the
key pre-exists; Copilot behavior is unchanged (API_KEY only)
- Pin the previously uncovered cases with tests: codex without a
pre-existing key, URL-only codex, and Copilot removing a stale
AUTH_TOKEN
---------
Co-authored-by: codeasier <liuyekang@huawei.com>
Co-authored-by: Jason <farion1231@gmail.com>
Insert a new 'claude-mythos-5' model tuple into src-tauri/src/database/schema.rs. The tuple ("claude-mythos-5", "Claude Mythos 5", "10", "50", "1.00", "12.50") is added to the models list (placed before the Claude 4.8 series) to register the Mythos 5 model with the Database schema.
Changing app_config_dir relocates the SQLite database, so a restart
triggered while proxy takeover is active used to strand the live
configs: the new instance reads a fresh DB with no live backups, the
first-run import then persisted the PROXY_MANAGED placeholder as the
`default` provider, and the no-backup recovery path wrote that
placeholder right back to the live files — leaving Claude/Codex/Gemini
pointed at a dead local proxy with no automatic way out.
Three orthogonal fixes, defense in depth:
- restart_app now awaits cleanup_before_exit() before app.restart().
Since #4069 the ExitRequested handler intentionally defers restart
requests to Tauri's default re-exec without custom cleanup, which is
correct for same-DB restarts but not for this command's dir-change
use case: only the old instance holds the backups needed to restore
the taken-over live files, so it must restore them while its event
loop is still alive.
- import_default_config refuses to import a live config that is under
proxy takeover (placeholder detected), instead of persisting it as
the current provider.
- restore_live_from_ssot_for_app validates that the current provider's
settings_config does not itself contain takeover placeholders before
writing it back; polluted SSOT now falls through to the placeholder
cleanup fallback.
Regression tests cover the import guard and the no-backup recovery
path (the latter fails before this change by writing PROXY_MANAGED
back to live).
* fix(updater): drive download/install/restart from backend to avoid hang
The 3.16/3.16.1 update flow was frontend-driven: downloadAndInstall()
then relaunchApp(). relaunch() routed through AppHandle::restart(), and
the old WebView had to keep running JS after the .app bundle was already
swapped — an unstable window that could hang the update or leave the old
version running until a manual restart.
Move the whole download -> install -> cleanup -> restart chain into a new
backend command install_update_and_restart, so it no longer depends on the
old WebView running JS after the bundle is swapped.
Platform-aware install ordering (install() behaves differently per OS):
- Windows: install() launches the external installer and exits the process
internally, so cleanup + single-instance destroy must run before install.
Surface a recovery hint on failure since the proxy may already be stopped.
- macOS/Linux: install() returns, so install first then cleanup — an install
failure no longer wrongly stops the proxy / reverts takeover.
Eliminate the restart vs single-instance race: restart_process() destroys
the single-instance lock (remove socket on macOS, ReleaseMutex on Windows)
before tauri::process::restart(), so the freshly spawned process can't
connect to the old listener and exit itself.
Also remove the now-dead frontend update plumbing (relaunchApp,
UpdateHandle, mapUpdateHandle) and surface backend errors in the toast.
Adapted from the original af4271f4 while rebasing onto #4069: the
ExitRequested handler changes were dropped entirely — the classifier from
#4069 already routes RESTART_EXIT_CODE to Tauri's default restart flow,
and the original should_restart branch (prevent_exit + async cleanup)
would have reintroduced the window-state deadlock that #4069 fixed.
install_update_and_restart bypasses ExitRequested entirely, so the two
fixes compose cleanly.
* fix(updater): clear tray icon on the direct-restart update path
restart_process re-execs via tauri::process::restart (spawn + exit(0)),
which skips Tauri's internal cleanup_before_exit and all RunEvent::Exit
plugin hooks. Window state, proxy/live restore and the single-instance
lock were already compensated explicitly; the tray icon was not.
On macOS/Linux the OS drops the status item when the process dies, so
the gap there was cosmetic at most. The real residue risk is the
Windows branch, which never reaches restart_process at all:
update.install() exits the process inside the updater plugin
(std::process::exit(0)), bypassing TrayIcon::drop — no NIM_DELETE is
sent and a stale icon lingers in the shell until hovered, the same
failure remove_tray_icon_before_exit was originally added for on the
quit path.
Call remove_tray_icon_before_exit (set_visible(false), proxied to the
main thread via run_item_main_thread) in restart_process and before
the Windows install. Deliberately not AppHandle::cleanup_before_exit():
it drops tray icons on the calling thread, which is not safe off the
main thread on macOS (NSStatusItem).
The updater's relaunch() (and app.restart()) triggers ExitRequested
with code RESTART_EXIT_CODE, which the handler treated as a regular
exit: it called api.prevent_exit() and spawned an async cleanup task.
However Tauri silently ignores prevent_exit() for restart requests
(see ExitRequestApi::prevent_exit docs), so the event loop keeps
shutting down regardless and fires every plugin's RunEvent::Exit hook.
Two threads then deadlock:
- the spawned cleanup task runs save_window_state on a tokio worker,
holding the window-state plugin's internal mutex while querying
window geometry, which dispatches to the main thread and blocks;
- the main thread, already inside the plugin's own RunEvent::Exit
hook, blocks on that same mutex.
The app freezes forever on the restarting screen with the update
already installed; force-quit + reopen comes back on the new version
(#3998). Confirmed on macOS by sampling the frozen process: main
thread parked in tauri_plugin_window_state save_window_state mutex
lock, tokio worker parked in is_maximized -> mpsc recv.
Fix: classify exit requests (None / restart / user exit) and let
restart requests fall through untouched to Tauri's default restart
flow (RunEvent::Exit -> re-exec). On that path window state is saved
by the plugin's exit hook on the main thread, the tray icon is
cleaned up by Tauri's internal cleanup_before_exit, and proxy/live
config restore is unnecessary because the new instance takes over
immediately. The regular exit path (tray quit) is unchanged.
With the fix, a simulated updater relaunch (request_restart) re-execs
the new process in under 200ms, 3/3 runs; normal quit still performs
full cleanup.
Co-authored-by: thisTom <19346741+thisTom@users.noreply.github.com>
The Claude/Codex format-transform non-stream branch returned an opaque 422
"Failed to parse upstream response" whenever a 2xx upstream body was not
valid JSON. The common case: MaaS gateways force-stream a stream:false
request and return an SSE body with a non-SSE Content-Type, defeating the
header-only is_sse() check.
On serde failure, sniff for SSE and aggregate the chunks into a single
JSON, then run the existing converter so clients still receive a valid
non-stream response.
- chat_sse_to_response_value: aggregate chat.completion.chunk SSE
(content / reasoning / refusal / tool_calls / legacy function_call),
tool_calls index-keyed via BTreeMap to avoid unbounded densification,
first-wins finish_reason, message-snapshot override, completeness and
error-event guards; synthesize an id when the upstream omits one
- responses_sse_to_response_value: process the residual trailing block,
tolerating truncation and skipping it once a completed event was seen
- enrich remaining parse failures with content-type / content-encoding /
body-snippet diagnostics
- deflate: try zlib (RFC 9110) before raw; keep the content-encoding
header for unsupported encodings
- gate zero-usage rows on the Claude transform path
Extract a shared `parse_custom_user_agent` helper in provider.rs returning
`Result<Option<HeaderValue>>`, and reuse it in the forwarder, stream check,
and model fetch paths so detection, forwarding, and model listing all apply
the same provider-level User-Agent. Previously only the forwarder honored it,
so stream check could fail (or model listing 403) on UA-gated upstreams that
the proxy itself handled fine.
- stream_check injects the provider's custom UA on the claude/codex paths and
still skips the GitHub Copilot fingerprint UA.
- model_fetch service + command and the model-fetch.ts wrapper thread an
optional UA through to GET /v1/models.
- runtime callers silently ignore invalid values via `.ok().flatten()`
(no save-time block, so deeplink imports stay lenient).
The model mapped for takeover (env mapping, Claude Desktop routes,
Copilot normalization, Codex chat override) was discarded inside the
forwarder, so usage attribution depended entirely on the upstream
echoing it back. When the upstream omitted the model or mirrored the
client alias, kimi/glm tokens were recorded and priced as claude-*
(roughly 5-25x overstatement).
- capture the final outbound model in forward(), return it via
ForwardResult, and store it on the request context
- attribution fallback order is now: upstream echo (empty string
treated as missing) -> outbound model -> client-requested model
- 'request' pricing mode anchors to the outbound model instead of the
pre-mapping client alias; unchanged when no mapping applies
- persist the resolved pricing_model on every usage row
- Claude Desktop rows now log app_type "claude-desktop" on streaming
and transform paths too (was hardcoded "claude", silently dropping
desktop provider pricing overrides and splitting the cost basis by
the stream flag); its global pricing defaults inherit the claude
config since proxy_config only allows claude/codex/gemini rows
- proxy_request_logs: add pricing_model column recording the basis actually
used at write time (NULL = pre-v11 rows, '' = unpriced error rows)
- cost backfill recomputes strictly by the persisted basis; the
request_model fallback now only applies to placeholder models, so
real-but-unpriced takeover rows stay at zero cost until pricing is
added instead of being permanently frozen at the alias's price
- backfill_missing_usage_costs_for_model can locate rows by pricing_model
- usage_daily_rollups: rebuild with request_model + pricing_model in the
primary key so the alias-to-real-model mapping and the pricing basis
survive the 30-day prune; legacy rows migrate with ''
- rollup_and_prune backfills costs before pruning: prune is irreversible
and used to run before the startup backfill, permanently booking
then-unpriced rows as zero
- get_model_stats groups by the effective pricing model
(COALESCE(NULLIF(pricing_model,''), model)) so costs aggregate under
the model whose prices produced them; response-mode behavior unchanged
The Zhipu quota API returns two TOKENS_LIMIT entries whose identity was
inferred by sorting nextResetTime ascending (nearest = five_hour). In the
last hours of each weekly cycle the weekly window resets sooner than the
current 5-hour session window, so the two buckets were swapped exactly
when users check their weekly quota most.
Classify by the explicit unit field instead (3 = hour window -> five_hour,
6 = week window -> weekly_limit; same shape on bigmodel.cn and api.z.ai,
weekly observed with number 7 and 1 so only unit is matched), falling back
to the old reset-time heuristic when the field is missing.
Codex /responses requests routed to text-only OpenAI-chat upstreams
(e.g. DeepSeek deepseek-v4-flash) failed with HTTP 400 "unknown variant
image_url" when images were sent: the responses->chat conversion turns
input_image items into image_url blocks the model rejects. The media
rectifier previously covered only the Claude adapter, so neither the
proactive strip nor the reactive retry fired for Codex.
- media_retry_should_trigger: accept "Codex" adapter, not just "Claude"
- contains_image_blocks / replace_images: also scan responses `input`
(input_image) in addition to chat `messages`
- is_image_block_type: match image | image_url | input_image
- is_unsupported_image_error: add "unknown variant" hint for the
deserialize error
- forward(): proactively run apply_media_prevention for Codex after the
responses->chat conversion
Proactively strips images for known text-only models (heuristic on by
default) and reactively retries with images replaced on upstream
image-unsupported errors. Adds tests for chat image_url, codex
input_image, the reactive trigger, and the deserialize error match.
Builds on #2774 (which fixed cache_read for the streaming openai_chat path).
Two gaps remained, both double-counting cache tokens when a Claude client
meters as app_type="claude" (input_includes_cache_read=false):
1. cache_read was still added to input on the non-streaming openai_chat path
(transform.rs openai_to_anthropic) and the whole openai_responses family
(transform_responses.rs build_anthropic_usage_from_responses, covering the
non-streaming call site and both streaming_responses call sites).
2. cache_creation was never subtracted on any converted path, including the
streaming openai_chat path #2774 had already touched. Claude billing treats
cache_creation as a separate bucket, so an inclusive upstream carrying a
direct cache_creation_input_tokens field billed it twice.
All four metering points now compute:
input = prompt_tokens - cache_read - cache_creation
restoring the invariant input + cache_read + cache_creation == prompt_tokens.
Pure OpenAI upstreams are unaffected (no cache_creation concept/field).
Tests: update direct-cache assertions (40->20), add a streaming conservation
regression test, and pin prompt<cache underflow (saturating clamp to 0) for all
three metering functions. cargo test 1573 pass, clippy clean.
Note: fix is forward-only; historical rows are not recomputed (cost is frozen at
log time and app_type="claude" mixes native + converted rows).
Audited all proxy format-conversion paths (Chat<->Message, Chat<->Response,
Gemini<->Message) for usage/cache metering. Five issues found and fixed.
The dedup mechanism (request_id PK, proxy/session source isolation) is
untouched, so no double-counting is introduced.
- A (Claude + openai_chat, streaming): inject stream_options.include_usage
so OpenAI-compatible upstreams emit usage in the SSE tail. Without it the
converted Anthropic message_delta was all-zero and the whole request's
input/output/cache was dropped. Same root cause as the already-fixed
Codex Chat path; the injection is extracted into a shared helper
(transform::inject_openai_stream_include_usage) reused by both paths.
- C (Claude + gemini_native): subtract cachedContentTokenCount from
input_tokens in build_anthropic_usage so input becomes fresh input
(Anthropic semantics). Previously the cache-hit tokens were billed twice
because this path meters as app_type="claude" (input_includes_cache_read
= false) while Gemini's promptTokenCount includes the cache.
- D (Codex + openai_chat, streaming): gate log_usage on
has_billable_tokens() to skip the synthetic all-zero usage the converter
emits when a non-compliant upstream omits usage, preventing empty-row
request-count inflation.
- P2 (from_claude_stream_events): use has_billable_tokens() for the return
gate instead of input>0||output>0, so a fully-cached streamed request
(cache_read>0, input==output==0) is still recorded. Affects all
Claude-streaming paths, not just Gemini.
- P3 (Codex Chat->Responses, non-streaming): apply the same
has_billable_tokens() filter the streaming branch got, since the
synthesized all-zero usage makes from_codex_response return Some and
bypass the `if let Some` guard.
Add TokenUsage::has_billable_tokens() as the unified predicate. New tests
cover include_usage injection, gemini input subtraction, the gate itself,
cache-only stream recording, and synthetic all-zero codex usage.
Full lib suite: 1569 passed.
The local session-log scanner dropped any assistant message that lacked
a stop_reason or had output_tokens==0. Claude Code Workflow / sub-agent
fan-out frequently produces messages that only wrote a message_start
snapshot (output=1, stop_reason=None) without a final block, yet their
input + cache_read + cache_creation tokens are already billed by
Anthropic (charged once the request is accepted). Dropping them
under-counted usage by ~4.1% overall, 92% concentrated in
workflow/subagent transcripts.
Replace the stop_reason/output gate with a billable-token check (any of
input/output/cache_read/cache_creation > 0). The per-message-id dedup
selection is unchanged, and request_id = "session:"+msg_id PRIMARY KEY
with INSERT OR IGNORE keeps each message single-inserted, so relaxing
the gate cannot double-count. Add a regression test covering a
stop_reason-less message with real cache cost plus an all-zero skip.
This is the parser-layer half of the Workflow under-counting fixed at
the collector layer in 8d332925.
collect_jsonl_files only walked <project>/<session>/subagents/*.jsonl,
so it missed Workflow sub-agent transcripts which live one level deeper
at subagents/workflows/wf_*/agent-*.jsonl. As a result all Workflow
token usage was invisible to the no-proxy session-log accounting.
Descend into subagents/workflows/wf_*/ as well, via a new
push_jsonl_children helper that keeps the fixed-depth, no-recursion
design. journal.jsonl carries no assistant rows so it is skipped at
parse time and needs no filename special-casing. Existing dedup
(request_id PK + INSERT OR IGNORE + should_skip_session_insert) keeps
the next sync's backfill idempotent.
Add test_collect_jsonl_files_includes_workflow_subagents.
The JS-script usage path resolved {{apiKey}}/{{baseUrl}} with env-only
field guessing, so apps that store credentials elsewhere (Codex:
auth.OPENAI_API_KEY + config.toml base_url) always got empty values and
custom-template queries failed despite a fully configured provider.
- query_usage / test_usage_script now delegate to
Provider::resolve_usage_credentials, the same per-app resolver used by
the native balance/coding-plan path and mirrored by the frontend
getProviderCredentials; explicit non-empty script values still win
- test_usage_script loads the provider and applies the same fallback,
so testing matches what a saved script does
- the custom-template variable preview shows the effective values
(script overrides first, then provider config) instead of always
showing provider credentials
- extract_codex_base_url documents and test-locks the frontend-mirror
invariant: non-active [model_providers.*] sections are never read
Reworked from the original patch to reuse the existing resolver instead
of duplicating per-app extraction.
Co-authored-by: Jason <farion1231@gmail.com>
* fix: prevent duplicate YAML keys in Hermes config
Three changes in hermes_config.rs:
1. deduplicate_top_level_keys() - scan and remove duplicate top-level
keys before YAML parsing, preventing "duplicate entry" parse errors
2. remove_all_sections() - helper to strip all occurrences of a given
top-level key from raw YAML text
3. replace_yaml_section() now calls remove_all_sections() on the
remainder after replacing the primary occurrence, preventing
duplicate sections from accumulating on repeated writes
Fixes the issue where mcp_servers (or any top-level key) gets
duplicated in config.yaml, causing "Failed to parse Hermes config
as YAML: duplicate entry with key" errors.
Co-Authored-By: que3sui <204201112+que3sui@users.noreply.github.com>
* fix: handle CRLF and LF line endings in top-level key deduplication
is_top_level_key_line only accepted empty, space, or tab after the colon,
but deduplicate_top_level_keys uses split_inclusive('\n'), so lines end
with \n (LF) or \r\n (CRLF). Without accepting \r and \n as valid
post-colon characters, the dedup safety net never activates.
Add \r and \n checks to is_top_level_key_line, and three tests covering
LF, CRLF, and first-occurrence preservation.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* refactor(hermes): keep last occurrence when healing duplicate YAML keys
Reworks the healing layers on top of the CRLF root-cause fix:
- deduplicate_top_level_keys: keep the LAST occurrence of each duplicated
key instead of the first. Duplicates come from section replacement
degrading into appends (#3633), so the last block is the newest data --
and Hermes itself reads the config with PyYAML, whose duplicate-key
semantics are last-wins. Keeping the first occurrence would silently
roll users back to stale config and diverge from what Hermes runs with.
Healthy files take a fast path and are returned untouched.
- Drop the unused dup_key variable (fails cargo clippy -- -D warnings,
which CI enforces).
- replace_yaml_section: clean residual duplicate sections from the
remainder via remove_all_sections; values come from the keep-last
healed read, so dropping all stale on-disk copies loses nothing.
- Add regression tests for the actual root cause (find/replace on CRLF
input must replace in place, not append), keep-last semantics,
identity on healthy files, end-to-end heal-then-parse, and duplicate
cleanup on write.
Fixes#3633#2973#2529#3310#3762
---------
Co-authored-by: que3sui <204201112+que3sui@users.noreply.github.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: Jason <farion1231@gmail.com>
* fix(proxy): strip cache_control from OpenAI format conversion (#3805)
- Remove cache_control passthrough from system messages, text blocks,
and tools to prevent 400 errors on strict OpenAI-compatible endpoints
- Always simplify single text block content to plain string format
- Fixes two format conversion bugs reported in issue #3805
* fix(proxy): apply cargo fmt to fix CI formatting check
Convert Responses input_file (requiring file_id or file_data, never file_url which Chat file parts do not support) and input_audio parts into their Chat Completions equivalents, and handle top-level input_* items that previously fell through and were dropped, clearing stale pending reasoning for non-assistant messages.
Replace the unconditional finalize at chat-to-responses stream end with a three-way guard: complete normally when finish_reason or [DONE] arrived, emit an incomplete response when substantive output exists without a finish_reason, and emit a failed (stream_truncated) event for empty truncation instead of masking it as completed. Also propagate late-arriving reasoning_content onto still-active tool-call items.
Generalize the cross-turn reasoning cache in codex chat history from function_call only to the full tool-call triad (function_call, custom_tool_call, tool_search_call) and their *_output counterparts, so apply_patch and tool-search calls keep their reasoning_content when restored via previous_response_id.
When listen_port is 0 the OS assigns the port at bind time, so the
configured value can no longer be trusted for building takeover URLs.
- server: read listener.local_addr() after bind and propagate the
actual port to the global proxy port, status, and ProxyServerInfo
- services: start the proxy before takeover when port is 0 so live
configs get the real port instead of :0, and persist the resolved
port back to the DB for DB-only URL paths; stop the pre-started
server on any takeover failure
- claude_desktop: reject an unresolved :0 port instead of emitting a
broken gateway URL
- build_proxy_urls: prefer the running server's port and error out if
the port is still 0
Add tests for takeover with an ephemeral port and the claude_desktop
:0 rejection; switch existing codex takeover tests to an ephemeral
port for isolation.
* feat(proxy): add GET /v1/models endpoint for Codex CLI reachability check
Codex CLI probes GET /v1/models at startup. Without this endpoint the proxy
returns 404, causing Codex to fail before any request reaches the upstream
LLM.
Return an OpenAI-compatible model list derived from the cc-switch–managed
model catalog file.
Fixes#3812
* fix(proxy): return Codex catalog schema from /v1/models
Codex deserializes the response as a catalog with a top-level `models`
field, not the OpenAI `{"object":"list","data":[...]}` envelope.
Return the catalog file content directly so the format matches what
Codex expects.
Co-authored-by: Codex review bot
* fix(proxy): guard /v1/models against serving stale catalog
Only return the model catalog when config.toml still references it via
`model_catalog_json`. After switching to a provider without a custom
catalog, the old file lingers on disk — serving it unconditionally
would advertise the previous provider's models to Codex.
Co-authored-by: Codex review bot
* fix(proxy): match relative model_catalog_json in stale-guard
cc-switch writes `model_catalog_json = "cc-switch-model-catalog.json"`
(relative) via set_codex_model_catalog_json_field. Match on the
filename constant rather than the absolute path so the guard works
with both relative and absolute paths.
Co-authored-by: Codex review bot
* fix(proxy): parse model_catalog_json field instead of substring match
Replace raw config_text.contains() with proper TOML field parsing so
commented-out lines and stray mentions of the filename in other fields
don't defeat the stale guard. Also switch from contains() to exact
filename match (Path::new(val).file_name() == Some(...)) to stay
consistent with resolve_cc_switch_catalog_path in codex_config.rs.
Add log::debug! when the guard blocks serving so the operator can
distinguish "no models configured" from "guard blocked stale catalog".
* refactor(proxy): reuse resolve_cc_switch_catalog_path in handle_models
Replace the inline config.toml parsing and filename match in
handle_models with the existing resolve_cc_switch_catalog_path helper
(now pub(crate)). This removes the duplicated stale-guard logic, keeps
a single source of truth for catalog-path ownership, and makes the
handler honor absolute model_catalog_json paths the same way Codex
live-setting import does.
---------
Co-authored-by: Jason <farion1231@gmail.com>
SaladDay