* fix(security): harden GrokBuild credential handling and Codex/Anthropic transforms against malformed input
Three robustness/security fixes in the upstream v3.18.0 code, each with a
regression test.
1) grok_config: remove the unconditional XAI_API_KEY fallback in
extract_credentials(). Credentials now come only from an explicit inline
api_key or the process env var named by env_key. Silently substituting a
different account's key (when the declared env_key var is unset) could
leak that key to whatever base_url the config points at.
2) deeplink/provider: merge_grokbuild_config no longer resolves env vars
into a plaintext api_key on import. A deeplink is untrusted input;
resolving+inlining would persist the victim's environment secret into the
imported provider's config.toml and ship it to the link's declared
base_url. env_key now stays an indirection (name), not a resolved secret.
3) proxy transforms: stop panicking on malformed upstream data.
- transform_codex_anthropic::anthropic_sse_to_message_value: only store a
`message`/`content_block` when it is an object; otherwise treat it as
empty, so the later `["content"]`/`["text"]`/`["signature"]` index
assignments can't panic on a scalar/array Value.
- streaming_codex_anthropic::responses_sse_events_from_anthropic_message:
bail out with a failed-event when the buffered body is a top-level JSON
array/scalar (a gateway that ignores stream:true), instead of
index-assigning into a non-object.
- mcp/grokbuild sync: normalize a non-table `mcp_servers` before
inserting, avoiding a toml_edit IndexMut panic on a user-edited
config.toml.
Panic findings verified with a minimal repro (serde_json index-assignment on
a non-object Value aborts). cargo test --lib: 2183 passed / 0 failed.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(mcp): normalize a non-table `mcp_servers` in Codex config before insert
Same class of bug as the GrokBuild MCP fix in the previous commit, but in the
much more widely used Codex path.
`sync_single_server_to_codex` guarded only with `contains_key("mcp_servers")`,
so a user-edited `~/.codex/config.toml` where the key exists but is *not* a
table (`mcp_servers = "x"` / `[]` / `42`) skipped the rebuild and then hit
`doc["mcp_servers"][id] = …`, which panics in toml_edit's `IndexMut`
(`.expect("index not found")`). The panic happens inside a Tauri command and
unwinds across the FFI boundary; in the provider-switch flow it also fires
after the DB/live write already committed, leaving a half-applied switch.
Extract the normalize-then-insert step into `upsert_mcp_server_table()` so the
previously panic-prone logic is unit-testable without touching the real
`~/.codex/config.toml`, and normalize any non-table value to an empty table
before inserting.
Audited the sibling MCP writers while here: claude.rs / gemini.rs /
hermes.rs / opencode.rs do not have this pattern (their `contains_key` uses are
read-only checks), and the other index-assignments in codex.rs operate on
freshly built `Table::new()` values, which are safe.
Tests: 2 new regression tests (malformed non-table values normalize and insert;
an existing valid table keeps its entries). cargo test --lib 2185 passed / 0
failed; cargo clippy --lib -D warnings clean.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(security): reject path-traversal entries when extracting a skill repo archive
`SkillService::download_and_extract` built the output path from the raw ZIP
entry name (`file.name()`), stripped only the leading `<repo>-<branch>/`
component, and then `dest.join(relative_path)` → `fs::File::create`. A crafted
entry such as `repo-main/../../../evil.sh` therefore escaped the destination
directory (zip-slip): arbitrary file write outside the temp extraction dir.
The archive is third-party controlled: it is downloaded from
`https://github.com/<owner>/<name>/archive/refs/heads/<branch>.zip`, and a
skill repo (`owner`/`name`) can be added through an untrusted `ccswitch://`
deeplink (`deeplink/skill.rs`), so the attacker fully controls the archive
contents.
Fix: resolve each entry through `zip::read::ZipFile::enclosed_name()`, which
rejects `..` components and absolute paths, before stripping the archive's
root directory and joining onto `dest`. Unsafe entries are skipped with a
warning. This matches what the two sibling extractors in this codebase already
did correctly (`extract_local_zip`, `webdav_sync/archive.rs`) — this call site
was the one that had been missed.
The extraction loop is split out into `extract_repo_archive()` so the guard is
testable without network access.
Verified the test actually catches the bug: with the guard reverted to the old
`file.name()` behaviour the new test fails ("zip-slip entry must not escape
dest (temp root)"); with the guard in place it passes.
cargo test --lib: 2186 passed / 0 failed; cargo clippy --lib -D warnings clean.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(security): strip all credentials from the shared Gemini common-config snippet
`extract_gemini_common_config` skipped only two hardcoded keys
(`GOOGLE_GEMINI_BASE_URL`, `GEMINI_API_KEY`) and copied every other `env`
entry into the shared snippet. But `GOOGLE_API_KEY` is a first-class Gemini
credential — `provider.rs` resolves it via
`first_non_empty(env, &["GEMINI_API_KEY", "GOOGLE_API_KEY"])` — so it was
never stripped.
That snippet is not inert: `apply_common_config_to_settings` (live.rs)
deep-merges it into the `env` of *every other* Gemini provider that uses the
common config, and the snippet is auto-extracted on startup, on import, and on
switch. Net effect: account A's key gets written into provider B and sent to
B's `GOOGLE_GEMINI_BASE_URL`, which may be a third-party relay. Anything else
the user put in `env` (`GOOGLE_APPLICATION_CREDENTIALS`, a proxy
`*_AUTH_TOKEN`, …) leaked the same way.
Fix: also skip `Self::is_sensitive_config_key(key)`, reusing the pattern
matcher the Claude extractor already relies on for exactly this reason (its
comment notes a fixed enumeration "will always miss the next `*_API_KEY`").
`GOOGLE_API_KEY` matches its `_API_KEY` suffix rule. Shareable non-secret
config (e.g. `GEMINI_TIMEOUT_MS`) is preserved.
Verified the test catches the bug: run against the old two-key filter it fails
with "credential GOOGLE_API_KEY must not leak into the shared Gemini snippet";
with the fix it passes.
cargo test --lib: 2187 passed / 0 failed; cargo clippy --lib -D warnings clean.
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(security): validate skill directory from backups and sync-imported DB rows
The skill install pipeline sanitizes the install directory name, but two
entry points bypassed it entirely:
- restore_from_backup joined metadata.skill.directory (raw meta.json
content) into the SSOT dir with no validation, allowing a crafted
backup to copy attacker-controlled files outside the skills directory
and to persist the poisoned value into the database.
- Sync import (WebDAV/S3) loads the remote database dump verbatim, so a
malicious or compromised sync snapshot could plant a skills row whose
directory contains path traversal. Every later raw join then operated
on attacker-controlled paths — uninstall/remove_from_app would
remove_dir_all outside the managed dirs (arbitrary directory deletion),
and sync_to_app_dir would write/symlink outside them.
Add require_valid_directory() (built on the existing
sanitize_install_name) and enforce it at restore_from_backup,
sync_to_app_dir, remove_from_app, and uninstall. Regression tests cover
all three sinks plus the metadata path; each was verified against the
unguarded code first (disabling the guard makes the restore/uninstall
tests complete the traversal write/delete, failing the assertions).
Also fixes a pre-existing cargo fmt drift in mcp/grokbuild.rs.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(security): reject path separators in sanitize_install_name on all platforms
The previous components()-based check was platform-dependent: on
Linux/macOS a backslash is not a path separator, so "a\b" parsed as a
single Normal component and was accepted as a valid install name — the
same value becomes a nested path when synced or restored onto Windows.
CI caught this on the Linux runner (the Windows runner passed).
Reject both '/' and '\' explicitly so the validation is
platform-independent; the components() check stays as the second layer
for dot segments, roots, and prefixes.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(security): close the remaining skill-install attack chain
The zip-slip guard added earlier only covered half the problem, and the
repo coordinates that decide *which* archive gets downloaded were never
validated at all. Together those two gaps let an attacker choose both the
bytes and where they land.
1) zip-slip, second layer. `enclosed_name()` only guarantees the entry does
not escape the *archive's own* root, and it does not normalise the path —
`..` survives verbatim in the returned value. Stripping `root_name`
afterwards spends one level of that depth budget, so `repo-main/../evil`
still escaped `dest`. On Windows it is worse: `root_name` comes from
`split('/')` and may contain backslashes, which Windows treats as
separators, so one `strip_prefix` can eat N components. Re-check the
*actual* relative path for `ParentDir` right before `dest.join()`.
Verified by reverting the guard: the single-`..` case escapes without it.
(The existing test used two `..`, which `enclosed_name()` rejects on its
own — it covered the half that was already guarded.)
2) Repo coordinates are now validated. `download_repo` formats owner/name/
branch straight into
`https://github.com/{owner}/{name}/archive/refs/heads/{branch}.zip`,
and `deeplink/skill.rs` applied no validation whatsoever. URL parsing
resolves dot segments, so a branch of
`../../../releases/download/v1/evil` retargets the request at a *release
asset* — arbitrary attacker-uploaded bytes. That is what made (1)
reachable in practice: git itself will not let a tree contain `..`.
Note the trigger is below "install": repos are enabled by default, so
merely opening the Skills panel downloads and extracts.
`validate_repo_ref` whitelists all three fields. Branch names legally
contain `/` (`feature/x`), so the check is per segment rather than a
blanket separator ban; an empty branch keeps its existing "use the
default branch" sentinel meaning, same as `HEAD`. The main guard sits in
`download_repo` — the single convergence point for all four call paths —
because both `skill_repos` and `skills` can be overwritten wholesale by a
sync snapshot, which no insert-time check can prevent. Entry-point checks
(deeplink, `add_skill_repo`, agents lock) exist so bad values surface
immediately instead of silently failing on every later panel open.
`build_skill_doc_url` is covered too: it feeds `readme_url`, which the
frontend hands to `openExternal`.
3) Regressions from the previous commits in this branch, both fixed:
- `uninstall` ran the directory guard *before* `db.delete_skill`. That
method has exactly one call site and is not exposed as a command, so a
row with a dirty `directory` (pre-v3.11.0 installs, or anything
`import_from_apps` still creates today — it has no sanitiser) became
permanently undeletable from the UI. Now the guard skips the filesystem
work but still deletes the row.
- `sync_to_app` propagated with `?`, so one bad row aborted the entire
app's skill sync — and that runs on provider switch. Now it warns and
continues per entry.
- `require_valid_directory` returned `sanitize_install_name`'s normalised
value, which trims. A DB value of `" foo "` would then join as `"foo"`
and miss the real directory. It now validates without rewriting.
4) Guard applied to the sinks the earlier commits missed: `migrate_storage`
(rename + remove_dir_all), `update_skill` (delete + write attacker-named
paths), `import_from_apps` (both a sink and the source of dirty values —
`selection.directory` arrives raw over IPC), `resolve_uninstall_backup_source`
(copies any directory into the backup area, which the UI then lists),
`check_updates`, and `migrate_skills_to_ssot`.
5) Extraction limits. The archive bytes were fully attacker-controlled via
(2), and this path had no ceiling of any kind, unlike
`webdav_sync/archive.rs`. Added entry count, total extracted bytes, a
symlink-target cap (zip 2.4.2's `make_reader` does not truncate at the
declared `uncompressed_size`, so a symlink-flagged entry that inflates to
gigabytes was read straight into memory), per-directory charging (a
directory-only archive writes no content bytes but still consumes inodes),
and a download-body cap (`response.bytes()` buffered the whole archive
before any limit applied). Budget is charged on bytes actually read and
written, never on sizes declared in the archive header.
Symlink materialisation is charged to the same budget, and a target that
contains the link itself (`dir/link -> ..`) is now rejected: the existing
"must stay inside base" check passes for it, and the recursive copy then
re-copies its own output every level until PATH_MAX.
6) Temp directories are now RAII. `download_repo` and `extract_local_zip`
called `TempDir::keep()` immediately and relied on every exit path
remembering `remove_dir_all`. Several did not — including
`fetch_repo_skills`, the highest-frequency path of all. Returning the
guard makes the leak unrepresentable at the call sites.
* fix(config): make removals and edits survive user-authored config shapes
The previous commits hardened the *write* side of the MCP tables against a
non-table `mcp_servers`. The read/delete side kept using `as_table_mut`,
which returns None for an inline table (`mcp_servers = { foo = {...} }` —
valid TOML). Removal then silently did nothing: the UI reported success, the
entry stayed in the file, and Codex loaded it again on next start. That is
worse than the panic it mirrors, because users usually reach for the toggle
precisely when an MCP server is misbehaving. Both Codex and GrokBuild now use
`as_table_like_mut` on both sides.
Normalisation is no longer silent either. Replacing a user's hand-written
`mcp_servers = "x"` with an empty table destroys data, so all four sites that
do it now log a warning first.
`update_codex_toml_field` had the same asymmetry with a different symptom:
when `model_providers` or `[model_providers.<id>]` was an inline table,
`as_table_mut` returned None and execution fell through to the "write a
top-level field" fallback. The user's `base_url` edit landed at the wrong
level, Codex never read it, and nothing reported a problem.
`opencode_config` parses `~/.config/opencode/opencode.json` with json5 into a
bare Value and never checked the root's shape. A user file of `[]` or a
scalar made `set_provider`, `set_mcp_server` and `add_plugin` all panic on
index-assignment, inside a Tauri command, unwinding across the FFI boundary.
(A top-level `null` is fine — serde_json promotes it.) Rejecting a non-object
root at the single read site fixes all three. Rejecting rather than rebuilding
is deliberate: the file also holds the user's own `model` / `theme` settings,
so a silent rebuild would delete them. Same treatment for the `provider` and
`mcp` sections, whose non-object forms made writes silently no-op.
* fix(proxy): recover a malformed Anthropic content_block as text
Sanitising a non-object `content_block` to `{}` stops the panic but creates a
quieter failure: the final Responses conversion matches on the block's `type`
and silently drops anything it does not recognise, so a garbled block header
turned into a `completed` response with empty output. The client saw the model
say nothing, with no signal that data had been discarded.
The replacement now carries `type: "text"`. The deltas that follow a malformed
header are usually well-formed, so this recovers the common case; a tool-use
block still yields nothing, exactly as before. A warning is logged when the
substitution happens.
The regression test now asserts through `anthropic_response_to_responses`
rather than on the intermediate value — asserting on the intermediate alone
passes while the client still receives an empty response.
* fix(grokbuild): decouple base_url from credential resolution, reject env_key-only links
`resolve_usage_credentials` called `extract_credentials(...).unwrap_or_default()`,
so a missing credential blanked the base_url too — even though it is written
right there in the config. Removing the `XAI_API_KEY` fallback in the earlier
commit widened that considerably: a GUI process on macOS does not inherit the
shell environment, so an `env_key` that resolves fine in a terminal resolves to
nothing here. The result was a Base URL shown in the UI that differed from the
one actually used, `{{baseUrl}}` expanding to empty in usage scripts (turning
requests relative), and native balance queries reporting "API key is empty"
while hiding the real cause. The two values are now resolved independently, via
the existing `grok_config::extract_base_url`, matching how the Codex arm of the
same function already works.
A deeplink whose only credential is `env_key` is now rejected by name. It was
already unimportable — `build_grokbuild_settings` has no `env_key` slot — but it
failed with the generic "API key is required", which reads like a malformed link
and invites the obvious "just carry the name over" fix. Carrying it over is
exactly what must not happen: the forwarder and the usage query both resolve
`env_key` at request time, so the victim's environment secret would still reach
the link's `base_url`. Same leak, merely deferred. The message says so, and the
test sets the probe variable so that restoring the resolution turns it red.
* fix(security): scrub credentials already leaked into the Gemini shared snippet
Fixing the extractor only prevents new contamination. A Gemini snippet is never
re-extracted once it exists — startup auto-extract and post-import extraction
both require `snippet.is_none()`, and the on-switch rewrite only covers Claude
and Codex — so existing users keep injecting the leaked key into live config,
the proxy upstream, and the new-provider form, where another account's key is
plainly visible.
Removing it from the snippet alone would make things worse. Merge and strip
cancel out by *value equality*: on switch-away, `remove_common_config_from_settings`
deletes the injected keys using the snippet as its only record of what to look
for. Once the key is gone from the snippet, the residue left in live config is
backfilled verbatim into the victim provider's `settings_config`, turning a
transient leak into a permanent one. So the cleanup covers all four locations at
once, and step order is itself a safety property: every fallible step runs before
the irreversible one, and a failure returns an error so the next start retries
from an intact state.
Deletion is by key *and* value, never by key name alone, so a provider's own
same-named key with a different value survives. The `~/.gemini/.env` edit
preserves layout rather than round-tripping through
`parse_env_file`/`serialize_env_file`: that pair drops comments, blank lines and
unparseable lines, collapses duplicate keys and re-sorts the file. Acceptable
when re-projecting everything, destructive for a targeted removal the user never
asked for — in testing it reduced a six-line fixture to one line, taking the
user's own key with it.
An audit record is written before any provider row changes, listing key names and
affected provider ids but *no values*: `settings` is not in `SYNC_SKIP_TABLES`,
so it is uploaded by WebDAV/S3 sync, and these are precisely the credentials that
must be destroyed. Keeping values would trade one deletion for a plaintext copy
that spreads across devices, has no UI to reach it, and never expires. It is
written only when absent, so a partially-applied earlier run cannot overwrite the
original pre-cleanup state.
Consequence, intentional: some Gemini providers will now report a missing API key
and need one entered. That key was never theirs. (The victim's own value was
already overwritten at merge time and cannot be recovered — worth noting in the
release notes, along with a recommendation to rotate the leaked key.) The snippet
row is deleted rather than set to `{}` when nothing shareable remains, and the
`cleared` flag is deliberately not set — either would disable auto-extract
permanently and prevent the user's legitimate shared config from ever coming back.
The frontend snippet validator is aligned with the backend matcher. It had the
same two hardcoded key names, so a hand-edited snippet could put `GOOGLE_API_KEY`
straight back with no error.
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Jason <farion1231@gmail.com>
Bump the default model IDs across every preset file and the downstream
defaults that mirror them:
- claude-opus-4-8 / anthropic/claude-opus-4.8 / global.anthropic.claude-opus-4-8
-> claude-opus-5, covering all three naming forms
- gpt-5.5 and the bare gpt-5.6 -> gpt-5.6-sol
- gemini-3.5-flash -> gemini-3.6-flash
Add gemini-3.6-flash to the built-in pricing seed (1.50 in / 7.50 out /
0.15 cache read per million). The seed runs INSERT OR IGNORE on every
startup, so existing databases pick the row up without overriding prices
the user edited by hand.
Advance the Claude Desktop opus route ID in step with the frontend SSOT:
CURRENT_OPUS_ROUTE_ID becomes claude-opus-5 and claude-opus-4-8 takes over
the LEGACY slot, so route IDs stored in existing user configs still resolve
through is_compatible_opus_route_alias.
Also sync omo.ts recommendations, form placeholders, the SudoCode partner
blurb and all four locales, plus the preset assertions that pin these IDs.
- Original sort mode now shows official > prime partner > sponsors
(preset file order, aligned with the README sponsor table) >
non-sponsors sorted by display name
- Physically reorder sponsor entries in all 8 preset files to match
the README sponsor table order; add group marker comments
- Drop stale section comments in openclaw/hermes preset files
- Update ProviderPresetSelector tests for the new ordering
Grok Build previously reused the whole Codex preset list, leaking
cn_official providers and Codex default models into the Grok CLI form.
Replace the inline filter with a standalone, independently maintained
preset module (no data linkage to codexProviderPresets):
- drop cn_official direct providers and open-source-only hosts
(SiliconFlow/ModelScope/Novita/Nvidia/AtlasCloud/OpenCode Go) that
have no Grok models upstream
- keep aggregators and third-party relays with grok-4.5 as the default
model (x-ai/grok-4.5 for OpenRouter-style namespaced routers)
- move the Grok Official seed entry into the presets module
- add standalone integrity tests; retarget the form's chat-mapping test
since no openai_chat preset remains in the list
Add a "Grok Official" preset and seed (grokbuild-official) whose empty
config represents the official login state: no custom [model.*] tables
are written, so Grok CLI falls back to its built-in xAI OAuth login and
cc-switch never touches those credentials.
Backend:
- Seed entry in OFFICIAL_SEEDS plus ensure_grokbuild_official_provider
command for on-demand repair (the one-shot master seeding flag is
already set for existing databases).
- Split validation into syntax-only (empty allowed) for live reads,
writes and official snapshots, keeping the full custom-model shape
check for non-official provider writes and imports. Backup/restore
can now round-trip an official-state live file.
- Manual import (command layer only) recognizes an official-state live
config and imports it as the official entry set as current, matching
the Codex official-login import outcome. Startup auto-import keeps
rejecting official-state live so a deleted official entry is never
resurrected on launch: startup import only captures real user data
as "default" and never manufactures official entries.
- Manual import also ensures the official entry before importing
(claude-desktop precedent) and after a successful custom import, so
first-time users end up with default + official like other apps.
- Proxy takeover guards skip or reject official-state live configs in
all three takeover paths, consistent with the official-provider
takeover ban.
Frontend:
- Grok Official preset entry in the GrokBuild form: official category
hides connection fields and passes the raw config through untouched.
- Filter managed-OAuth presets out of the GrokBuild preset list; they
were never wired for this app and produced keyless broken configs.
Tests cover seed presence, official round-trip, ensure-after-deletion,
and the four import scenarios including startup non-resurrection.
Managed-OAuth providers (github_copilot / codex_oauth / xai_oauth) need
proxy takeover to inject credentials, but the "needs routing" badge and
the switch-time warning only keyed off apiFormat — missing OAuth providers
whose upstream format is native (e.g. the new Codex xAI Grok OAuth preset
on openai_responses).
- Add isOAuthProviderType / OAUTH_PROVIDER_TYPES as the SSOT for OAuth types
- Extract providerNeedsRouting() as the authoritative predicate: managed
OAuth always needs routing (the backend rejects these accounts' direct
connection) regardless of apiFormat or Claude Desktop mode
- ProviderCard badges (claude / codex / claude-desktop) and the switch
warning consume the shared predicate instead of raw apiFormat
- Gate the warning on per-app routing readiness: claude-desktop uses
isProxyRunning (backend takeover status has no such field), other apps
use isProxyTakeover — fixes false warnings on Desktop and missing
warnings when the proxy runs but the app isn't taken over
- Force proxy mode for all managed-OAuth providers in the Claude Desktop
form (lock the mode toggle), not only xai_oauth
- Add unit tests for providerNeedsRouting, the switch routing gate, and
Desktop OAuth preset enforcement
- i18n(zh/en/ja/zh-TW): add notifications.proxyReasonManagedOAuth
Add a managed "xAI (Grok) OAuth" Codex provider that routes through the
local proxy to api.x.ai via the shared Grok CLI OAuth identity, plus the
native-Responses compatibility layer that makes Codex 0.142+ traffic work
against xAI's strict upstream serde parser.
Provider:
- codex.rs: recognize the xai_oauth placeholder in extract_auth, hard-pin
the base URL to api.x.ai and the tool profile to native Responses
- forwarder.rs: treat xAI OAuth auth failures as non-retryable
- presets + ProviderForm/CodexFormFields: managed OAuth preset that hides
the api key/endpoint fields and derives the provider type across apps
Native Responses compatibility (gated on is_xai_oauth, so no other
provider is affected):
- transform_codex_responses_namespace: flatten Codex's private
namespace/plugin tool declarations into top-level function tools on the
request; restore the flat function_call names back to {name, namespace}
on the response (streaming and non-streaming) so the client matches its
own namespaced tool registry
- transform_codex_responses_xai_sanitize: strip the OpenAI-backend-private
fields xAI rejects (external_web_access, prompt_cache_retention,
safety_identifier, the additional_tools carrier, tool_search, ...) with
deterministic removals that keep the prompt-cache prefix stable
- wire both into the native passthrough after the request transform;
response restore runs in a dedicated handler so the generic passthrough
hot path is untouched
Ports the proven approach of sub2api's Grok Responses gateway. Verified
with a 4-round codex -> xAI OAuth workload: all tasks green, zero upstream
errors.
Add XaiOAuthSection (device-code login, account list, default selection)
backed by the generalized useManagedAuth hook, wire it into the Auth
Center and both Claude provider forms, and expose model fetching for
signed-in accounts.
- Accounts requiring re-auth stay visible in the account selector as
disabled items with an expired badge instead of silently vanishing.
- Auth status refetches periodically so a revoked refresh token surfaces
without reloading the panel.
- All strings ship in zh/en/ja/zh-TW; a locale coverage test pins every
required key in all four locales.
Codex always sends prompt_cache_key in its Responses requests, but the
Responses -> Chat Completions conversion dropped it, breaking session
cache affinity on upstreams that route by key (e.g. Kimi Coding).
- Re-inject prompt_cache_key after conversion in the forwarder: an
explicit client key wins, otherwise a client-provided session ID;
generated per-request UUIDs are never sent upstream.
- Provider-aware gating: "auto" enables only known-compatible upstreams
(api.openai.com, api.kimi.com/coding) because strict gateways reject
unknown fields with HTTP 400 (e.g. Fireworks); an advanced
Auto/Enabled/Disabled override is available on the Codex form in all
four locales.
- Kimi For Coding preset opts in explicitly.
Expose the top-level `model` key of config.toml as an editable field in
the Codex provider form, so users can switch to newly released models
(e.g. gpt-5.6) without waiting for a preset update — preset updates only
affect newly added providers, existing ones keep their saved TOML.
The field syncs bidirectionally with the TOML editor (mirroring the
base-URL pattern), suggests models from the mapping catalog plus the
provider's /models endpoint, and offers a one-click "add to mapping"
action when the value is outside a non-empty catalog. Hidden for
official providers.
Details:
- Save-time catalog sync now only backfills the first mapping row into
the top-level model when the field was left empty, so the explicit
field wins over the implicit row-0 sync
- Escape model names (and base_url) with TOML basic-string escaping and
strip control characters from field input: /models ids are remote
data and unescaped interpolation could inject config.toml lines
(e.g. a forged [mcp_servers.*] command)
- The strict model-line matcher now recognizes escaped output, empty
strings and single-quoted literals, keeping extract/set round-trips
stable; deliberately no key-only loose matching (it would misfire on
assignment-looking text inside multiline strings)
- Invalidate the fetched model list whenever the request identity
changes (base URL, full-URL toggle, API key, custom UA) and discard
in-flight responses via a sequence guard, so the dropdown never shows
a previous provider's models
- i18n for zh/en/ja/zh-TW
* feat(codex): support native Anthropic Messages protocol as upstream
Allow gateways that only expose the native Anthropic Messages protocol
(/v1/messages) to be used by Codex: the local proxy performs bidirectional
request/response/streaming conversion between Responses and Anthropic.
Backend:
- Add two conversion modules: transform_codex_anthropic / streaming_codex_anthropic
- codex.rs: add routing detection and auth: ANTHROPIC_AUTH_TOKEN→Bearer (default),
ANTHROPIC_API_KEY→x-api-key, mutually exclusive
- handlers.rs: add handle_codex_anthropic_to_responses_transform
- forwarder.rs: support optional Claude Code client fingerprint impersonation
(User-Agent / anthropic-beta / x-app / system prompt first-line injection)
and /responses→/v1/messages rewriting
- codex_config: the anthropic format reuses the NativeResponses profile to strip
custom tools
- ProviderMeta: add impersonateClaudeCode
Frontend:
- CodexApiFormat: add "anthropic"; the form adds auth field selection and an
impersonation toggle
- Add en/ja/zh/zh-TW copy
Robustness:
- Downgrade when tool history / forced tool_choice conflicts with extended
thinking, avoiding upstream 400s
- Emit cache_creation_input_tokens in usage and use saturating_add to guard
against overflow
- Append a unique suffix to non-streaming/streaming output-item ids to avoid
multi-segment text/thinking overwriting each other
* fix(codex): harden Anthropic bridge against empty text blocks and truncated streams
- Drop empty/whitespace-only text content blocks when rebuilding Anthropic
messages from Responses history; Anthropic 400s on empty text blocks (e.g. an
empty assistant text emitted alongside a tool_use), which broke follow-up and
tool-result requests. Also drop messages left without content.
- Do not report a truncated Anthropic stream as completed: when the SSE
connection ends before message_stop with no stop_reason, emit an incomplete
response if partial output exists, or a failed (stream_truncated) response
otherwise, mirroring the chat converter's EOF handling.
* fix(codex): resolve Anthropic-bridge review findings
Blocking:
1. Defer stripping the [1m] long-context marker until after catalog matching and model write-back, re-stripping it on the final Codex→Anthropic body and setting a flag to emit the context-1m-2025-08-07 beta header, so the marker is no longer lost or overridden by the default model.
2. Gate Anthropic thinking on the trailing turn only (via trailing_turn_allows_thinking) instead of scanning full history, so a Codex session that resends history each round no longer permanently loses thinking after the first tool call.
3. Inject 5m ephemeral cache_control on the Codex→Anthropic body by reusing cache_injector (handling the system string→array conversion), so system/tools/history are cached instead of re-sent at full price every round.
4. Add a shared base_url_is_full_endpoint helper (normalizing whitespace/query/fragment/trailing slash) used by both the Anthropic and Chat paths, so a base URL already ending in /v1/messages is treated as a full endpoint instead of double-appending to /v1/messages/v1/messages.
5. Align the catalog tool-profile predicate with the routing predicate so resolve_codex_catalog_tool_profile returns the Anthropic profile whenever the request converts to Anthropic, preventing freeform tools like apply_patch from being silently filtered by a ProxyChat catalog.
6. Explicitly disable native web_search for the Anthropic profile (including the no-catalog branch) via set_codex_native_web_search_field, so Codex no longer treats it as available while the transform silently drops it.
7. Only forward tool_choice when tools survive filtering, dropping it otherwise, to avoid a non-retryable 400 from upstream when tool_choice is sent with no tools.
8. Lower the fallback default to max_tokens=8192 (only when max_output_tokens is omitted) and clamp the thinking budget to max_tokens/2, disabling thinking below the 1024 floor, to avoid hard 400s on low-output-ceiling models/gateways.
Minor:
9. Centralize the Codex/OpenAI fingerprint-header denylist in is_codex_client_fingerprint_header so impersonating Claude Code uniformly drops originator/session_id/conversation_id/chatgpt-account-id/x-client-request-id/openai-* and the x-stainless-*/x-codex-* prefixes.
10. Retain content_block_start.input as start_input and fall back to it at block close when no input_json_delta arrived, so a gateway that carries the full tool input on the start event no longer yields empty tool arguments.
11. Extract a shared codex_responses_sse module as the single Responses SSE envelope builder that both the chat and anthropic streaming emitters delegate to, with byte-for-byte-unchanged wire output, so future event-format fixes touch one place instead of two.
* fix(codex): add per-provider max_output_tokens override for Codex→Anthropic path
Codex does not forward model_max_output_tokens in the request body,
causing the proxy to fall back to a conservative 8192 default. This
truncates long or thinking-heavy responses (stop_reason=max_tokens).
- Add maxOutputTokens field to ProviderMeta (Rust + TypeScript)
- Inject the value into the Anthropic request body before transform,
taking precedence over request-supplied and default values
- Add numeric input in Codex form fields (Anthropic format only)
- Add i18n entries for label, placeholder, and hint (en/zh/zh-TW/ja)
- Include roundtrip and omission unit tests for the new field
* fix(codex): harden Anthropic protocol bridge
* fix(codex): address Anthropic bridge review
* fix(codex): preserve flattened Anthropic inputs
* fix(codex): harden Anthropic recovery paths
* fix(ci): satisfy Rust clippy
---------
Co-authored-by: Jason <farion1231@gmail.com>
The fallback model (ANTHROPIC_MODEL) field was missing the 1M checkbox
that the role-specific model rows already have. Added the checkbox with
the same strip/set marker logic used by role rows.
Fixes#3679
updateTomlCommonConfigSnippet re-serialized the whole document through
smol-toml (parse -> deepMerge -> stringify): comments dropped, keys
reordered, and empty parent table headers synthesized -- the
long-standing "config.toml keeps getting reordered" symptom (audit C5,
introduced in 083e48bf).
Replace it with a backend command backed by the same
merge_toml_table_like / remove_toml_table_like used when writing live
configs, so the form preview and the live write share one merge
semantic and user formatting survives edit-time merges. The frontend
sync helper is deleted outright to keep the pattern from coming back.
Making the form operations async exposes them to interleaving, so a
result is discarded unless it is still current when it lands:
- a per-hook sequence number (last operation wins) covers rapid
toggle/save races where an earlier merge resolves after a later
removal and would flip the switch back;
- a config-baseline check covers the user hand-editing the TOML while
a merge is in flight -- the stale result must not clobber their edit.
The checkbox state self-heals via the existing inference effect.
Regression tests pin both by resolving a suspended merge after a newer
operation / an external edit, plus backend tests locking comment and
key-order preservation, scalar override, and value-matched removal.
* feat: add Claude subagent takeover config
* feat: add Claude subagent model field
* i18n: add Claude subagent model labels
* fix(proxy): preserve configured subagent model mapping
* fix(providers): exclude subagent model from Claude common config
* style: format rust code
* fix(i18n): rename "Write Common Config" checkbox to "Apply Common Config"
The original label "Write Common Config" (写入通用配置) was ambiguous
about data flow direction — it read as "write current config INTO the
common config" when the actual behavior is the reverse: merge the
saved common config snippet INTO this provider's config.
Rename to "Apply Common Config" across zh/zh-TW/en/ja, including all
hint/guide/notice references and component defaultValue fallbacks, to
make the action unambiguous. Also aligns the Gemini hint wording with
the claude/codex "when checked" phrasing.
* docs(ja): sync Japanese docs with renamed 'Apply Common Config' label
Codex review on #4829 pointed out the Japanese user manual and README
still referenced the old checkbox label after the rename.
- docs/user-manual/ja/2-providers/2.1-add.md: 共通設定を書き込み → 共通設定を適用
- README_JA.md: 共有設定を書き込む → 共有設定を適用
The "Get API Key" link (and partner promotion) below the API key input was
only wired for claude/codex/gemini/opencode. Claude Desktop uses a separate
form that never rendered it, and OpenClaw/Hermes were blocked by two gaps:
- useApiKeyLink whitelisted only 4 app ids, so the link was suppressed for
claude-desktop/openclaw/hermes even when a preset carried apiKeyUrl.
- useProviderCategory only parsed claude/codex/gemini/opencode preset ids, so
OpenClaw/Hermes category stayed undefined and the link condition
(cn_official/aggregator/third_party) never held.
Changes:
- ClaudeDesktopProviderForm: call useApiKeyLink and replace the bare API key
Input with the shared ApiKeySection.
- useApiKeyLink: add claude-desktop/openclaw/hermes to the whitelist and add
ClaudeDesktopProviderPreset to the PresetEntry union.
- useProviderCategory: parse and resolve openclaw/hermes preset categories.
- Hermes/OpenClaw form fields: don't let an "official" category disable the
API key input, since these apps have no OAuth-only official providers
(e.g. Hermes' Nous Research is "official" but still needs a user key).
Align the Codex provider form with Claude Code. Model mapping (the
catalog) is independent from route takeover: native Responses providers
(MiMo, Doubao, MiniMax) need it for direct connect with no proxy, while
Chat providers use the proxy regardless of any per-provider flag.
- Remove the "Needs Local Routing" toggle. It had no backend field and
only gated catalog/reasoning persistence, which is equivalent to
"is the mapping filled".
- Always show model mapping for non-official providers; persist when the
list is non-empty (backend already keys off modelCatalog.models).
- Gate reasoning visibility/persistence on Chat format, not the toggle.
- Mark the Chat upstream-format option "routing required" and refresh the
advanced-section hint (zh/en/ja/zh-TW); drop dead localRouting* keys.
- Move the model-mapping block above the custom User-Agent block.
fix(codex): preserve native-profile hidden catalog fields on edit
useCodexConfigState dropped supportsParallelToolCalls / inputModalities /
baseInstructions when loading a saved provider, so editing and saving a
native Responses provider lost parallel tools, image input and the
official base instructions from the generated catalog. Preserve them on
load (camelCase + snake_case) and compare them in the row sync. Add a
regression test.
Codex providers can now run in two modes per provider:
- Proxy-Chat (route takeover): apiFormat=openai_chat, existing Responses<->Chat conversion. Unchanged.
- Native-Responses (direct): apiFormat=openai_responses, no proxy. cc-switch generates ~/.codex/cc-switch-model-catalog.json so Codex shows the custom models and tools work without the freeform apply_patch (type=custom) tool that native gateways like MiMo reject; editing falls back to shell_command.
Catalog generation is keyed on apiFormat (CodexCatalogToolProfile), decoupled from the takeover toggle, so native providers persist a catalog without enabling local route mapping.
Codex's catalog parser requires base_instructions on every entry; the native template carries a neutral default and per-vendor official text overrides it (MiMo, MiniMax). Synthesized catalogs for Qwen/Doubao/LongCat use the neutral default.
Existing native providers must be re-saved once to regenerate a valid catalog (no DB migration).
Decouple the in-app promotion banner from isPartner so a preset can show a
promo phrase plus referral link without earning the paid-partner star.
- ApiKeySection: gate the banner on partnerPromotionKey alone; the star is
still driven solely by isPartner. This also un-suppresses the existing
MiniMax cn/en promos, which already carried copy without isPartner.
- OpenCode Go (claude/codex/opencode): point apiKeyUrl at the referral
link and add partnerPromotionKey "opencode_go" (no isPartner set).
- i18n: add the opencode_go promo string across zh/en/ja/zh-TW.
The Codex provider form tied Chat-format conversion and route takeover (model
mapping) to one toggle, so a provider serving a native Responses API could not
use model mapping without forcing Chat Completions conversion.
- Promote the upstream format (Chat Completions / Responses) to an independent,
always-visible selector that triggers no sub-menus on its own.
- The local-routing toggle now solely gates the advanced sub-sections: model
mapping catalog, plus reasoning capability when the format is Chat.
- Persist modelCatalog / codexChatReasoning based on the takeover toggle and
derive its initial state from saved catalog presence (no new persisted field).
- Refresh codexConfig i18n (zh/en/ja/zh-TW): add upstreamFormat* keys and reword
the routing/advanced hints to reflect the decoupling.
- Fix codexChatProviderPresets test expectation for the Doubao Seed 2.1 Pro rename.
In the default (Original) sort mode, float official-category presets to
the very top, followed by prime partners (Kimi), then the rest in their
original array order. Implemented as a three-way partition+concat so each
group keeps its internal order and no preset is listed twice.
Introduce an optional primePartner flag on all provider preset interfaces.
Regular partners keep the gold star badge; prime partners render a solid
gold heart (no badge frame). Mark the first-party Moonshot Kimi presets
(Kimi / Kimi For Coding / Kimi K2.7 Code) across claude, codex, claude
desktop, opencode, openclaw and hermes as prime partners.
* fix(provider-preset-selector): after searching presets, it is impossible to select search results.
feat(provider-preset-selector): add keyboard shortcut for search input and improve focus handling.
* fix(provider-list): prevent keydown event from triggering when default is prevented
* fix(provider-preset): keep preset clickable after search and restore keyboard UX
- ProviderList: scope the typing guard to the Ctrl/Cmd+F branch so Escape
still closes the search panel (the top-level early return swallowed it);
reuse isTextEditableTarget instead of re-implementing the check.
- ProviderPresetSelector: drop the rAF select() that raced with typing and
ate the first character (gateway -> ateway), and restore the input
autoFocus for the open-by-click path; refocus via rAF when Ctrl/Cmd+F is
pressed while the box is already open so focus returns to the input.
- Add a regression test for the re-focus-on-shortcut behavior.
---------
Co-authored-by: Jason <farion1231@gmail.com>
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.
* 调整预设供应商按钮外观与搜索框位置
1. 调整预设供应商按钮外观,显示默认图标,大小统一;
2. 调整预设供应商搜索框位置。
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* test(provider): 新增预设按钮外观与 inline 搜索的单元测试
覆盖:
1. 所有预设按钮固定 200px 宽度,视觉对齐一致
2. preset.icon 存在时按钮内渲染 ProviderIcon
3. preset 无 icon 且无 theme.icon 时渲染占位元素保持文字对齐
4. 点击放大镜 inline 切换搜索输入框可见性,ESC 收起并清空
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* refactor(provider-preset): responsive grid layout and search polish
- Replace fixed-width preset buttons with a responsive CSS grid (auto-fill, 150px min column)
- Add a leading placeholder to the custom button so its label aligns with iconed presets
- Close the inline search box on outside click, restoring the old Popover behavior
- Span the empty-state hint across the full grid row
- Update component tests for the new layout and behaviors
---------
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: Jason <farion1231@gmail.com>
The preset search text also included websiteUrl and the shared category label, producing imprecise matches: a single category term matched the whole group, and URL fragments like "com"/"api" matched nearly everything. Restrict the search text to the display name and raw name; category labels are still used for rendering.
The menu display name and request model columns used mismatched example
brands (DeepSeek vs Kimi), implying a display name maps to an unrelated
request model. Derive both placeholders from the row role so each row
stays brand-consistent, and route the lightweight Haiku tier to a flash
example (deepseek-v4-flash) while other tiers use deepseek-v4-pro.
- 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.
- Fold local routing toggle, model mapping, reasoning overrides and custom
User-Agent into a single collapsible advanced section, mirroring the
Claude form (auto-expands when UA is set or local routing is enabled)
- Custom User-Agent becomes configurable for native Responses providers;
it was previously reachable only when openai_chat routing was on
- Collapsed hint names local routing as the entry point for Chat
Completions / non-GPT providers
- Backfill all missing codexConfig keys in zh-TW locale
Polish the provider-level User-Agent override UI on the Claude and Codex forms.
- Add a shared CustomUserAgentField (label + input + preset dropdown + live
validation) so both forms stay in sync.
- Provide curated UA presets (Claude Code / Kilo Code families that pass
coding-plan UA whitelists per #3671); the first is Claude Code's real
`claude-cli/x (external, cli)` format. Whitelists gate on the name prefix,
not the version, so static values stay valid across upgrades.
- Expose presets via a dropdown to the right of the input (z-[200] so it
renders above the dialog layers) instead of inline chips.
- Move the field into the existing advanced/reasoning collapsibles.
- userAgent.ts mirrors the backend byte rule (reject only control chars;
non-ASCII is allowed) for a non-blocking inline hint.
- i18n for all four locales (zh/en/ja/zh-TW).
Restyle the proxy-takeover notice in the Codex editor from the boxed
Alert to the amber inline-hint style used by the endpoint hints, and drop
implementation jargon (127.0.0.1 / PROXY_MANAGED) that users could not
interpret. The notice and the auth/config hints now simply state that the
form shows the stored provider config rather than the proxy-managed live
config. i18n synced across en/zh/ja/zh-TW.
The reported "OAuth access token disappears when enabling Codex proxy
takeover" was a display artifact, not data loss: auth.json on disk kept
the OAuth login the whole time. During takeover the edit dialog falls
back to the stored provider config (so it does not surface the proxy
placeholder), which for a third-party provider shows that provider's own
key instead of the live auth.json, making the OAuth token look gone.
Thread an isProxyTakeover flag from App through ProviderForm into the
Codex editor and show an explicit notice plus storage-aware auth/config
hints clarifying that the form displays the stored provider config while
the live config is temporarily managed by the proxy. Drop the
proxy-running condition so the notice shows whenever takeover is active,
even with the proxy stopped.
Add a regression test asserting the dialog does not read live settings
during takeover and renders the database config. i18n synced across
en/zh/ja/zh-TW.
Claude Desktop's 3P validation only accepts claude-{sonnet,opus,haiku}-*
role IDs, so providers must map every tier. Bring the Desktop mapping flow
in line with Claude Code and fix the fallout that broke sub-agent Haiku calls.
- Proxy form: replace the dynamic route list with fixed Sonnet/Opus/Haiku
tiers; blank tiers backfill from the first filled tier (Sonnet first) on
submit and inherit its supports1m flag
- Backend: add a role-keyword fallback in map_proxy_request_model so dated
official names (e.g. claude-haiku-4-5-20251001) resolve to the right tier,
guarded by is_claude_safe_model_id so [1m]-suffixed IDs stay rejected
- Tighten is_claude_safe_model_id / isClaudeSafeRoute to reject degenerate
role IDs like "claude-sonnet-"
- Fix the seed-effect race where normalizing empty routes to three blank
tiers blocked the default-route backfill
- Sync switch hints, placeholders, and the zh/en/ja/zh-TW locales to the
three-role-ID rule
- Update zh/en/ja user manual (2.1, 2.6), calling out legacy Claude IDs
(claude-3-5-sonnet-...) as a rejected example
Tests: 282 frontend + 34 backend claude_desktop; typecheck, clippy, fmt clean.
Bump default model names project-wide: gpt-5.4 -> gpt-5.5,
gemini-3.x -> gemini-3.5-flash, glm-5 -> glm-5.1, and
grok-code-fast-1 -> grok-build-0.1 across all provider presets
(claude, codex, gemini, hermes, openclaw, opencode, universal),
Gemini config, and stream check defaults.
Pricing:
- Seed gemini-3.5-flash, gemini-3.1-flash-lite, step-3.5-flash-2603,
doubao-seed-2.0-code, mimo-v2.5(/pro), qwen3-coder-480b, grok-build-0.1.
- Correct deepseek-v4-flash/pro, glm-5/5.1, grok pricing.
- Add repair_current_model_pricing: idempotent pass that fixes only
rows still equal to the outdated built-in values, preserving any
user-customized prices (seed uses INSERT OR IGNORE and cannot update
existing rows).
Fixes from review:
- opencode: drop duplicate gemini-3.5-flash variant (unreachable via
.find), keep the entry with the full minimal/low/medium/high set.
- Align stale display names/costs to gemini-3.5-flash (hermes, openclaw,
opencode); openclaw cost -> {1.5, 9, 0.15} to match seed.
- i18n (zh/en/ja/zh-TW): refresh OMO category tooltips for new model
names; fix writing tooltip to Kimi K2.5 to match its recommended.
Update tests accordingly and add a regression test asserting unique
model ids in the Google opencode preset variants.
Bump the default Opus route/model from claude-opus-4-7 to claude-opus-4-8
across provider presets (claude, claudeDesktop, hermes, openclaw, opencode,
universal), i18n locales (zh/en/ja/zh-TW), pricing seed data, and the
user-manual docs.
- Add claude-opus-4-8 pricing row ($5/$25/$0.50/$6.25); keep the 4-7 row
for historical usage stats (seeded via INSERT OR IGNORE).
- Claude Desktop proxy: accept bidirectional opus 4-7 <-> 4-8 route alias
during rollout so previously saved routes keep resolving.
- thinking_optimizer: route opus-4-8 through adaptive thinking and normalize
dotted model ids (also fixes dotted 4-6/4-7 falling back to legacy).
- usage_stats: normalize Bedrock/Vertex/aggregator opus-4-8 ids to base
pricing.
Also merge role:"system" messages into the Gemini systemInstruction in the
Anthropic->Gemini transform.
The Fill Recommended button was misleading — it showed toast.success even
when most slots couldn't be filled due to model ID mismatches with the
user's configured providers.
Changes:
- Sync OMO_BUILTIN_AGENTS/CATEGORIES recommended fields with upstream
oh-my-openagent model-requirements (gpt-5.4→5.5, kimi-k2.5→claude-sonnet-4-6, etc.)
- Add toast.warning tier when unmatched >= filled, showing slot:model pairs
(e.g. "Sisyphus: claude-opus-4-7") so users know exactly what to configure
- Upgrade fillRecommendedNoMatch to also show examples
- Add fillRecommendedMostlyUnmatched i18n key (zh/en/ja/zh-TW)
zayoka