Innokentiy Solntsev bc7f68fb91 fix(resilience): lock the exact model, not the quota family, on 5xx model-lockout failures (#12957) ...

* fix(resilience): lock the exact model, not the quota family, on 5xx model-lockout failures

A 5xx model-lockout failure — a transport error (terminated, EHOSTUNREACH,
connect timeout), an upstream server error, or OmniRoute's own synthesized
502 from quality validation — is evidence about one model endpoint at that
moment, not about the account's quota family. recordModelLockoutFailure()
wrote it under the quota-family key regardless, so for codex (whose family
key is the whole `codex` scope, i.e. every gpt-5* model) one empty stream
on gpt-5.6-luna removed gpt-5.6-sol and gpt-5.6-terra from routing too,
for 2–30 min with exponential escalation, while the quota was untouched.

- exactModelLock.ts: resolveLockoutScope(status, explicit) — 429/403/402
  (and 404, already narrowed by getModelLockKey) keep the family key; any
  other status uses the exact provider/connection/model key. An explicit
  `scope` option still wins.
- recordModelLockoutFailure() resolves the scope once for key + lock fn.
- decayModelFailureCount() now walks every key shape (family, not_found,
  exact) so success-decay reaches exact-scope locks; null model stays a
  no-op.
- getAllModelLockouts() parses the `exact:` marker out of the key so the
  Model Cooldowns card lists the bare model and can clear it by that name.
- docs: RESILIENCE_GUIDE §3 key-scope-by-status; changelog fragment.

* chore(changelog): name the fragment after PR #12957 and link issue #12955

---------

Co-authored-by: insoln <is@careerum.com>

2026-09-17 16:25:32 -03:00

..

__tests__

fix(routing): table pricing with catalog fallback for off-table models, pooled latency bootstrap, fresh tier cache (#12792)

2026-09-10 09:20:18 -03:00

accountFallback

fix(resilience): lock the exact model, not the quota family, on 5xx model-lockout failures (#12957)

2026-09-17 16:25:32 -03:00

admission

fix(admission): measure request bodies with the active cost budget (#13762)

2026-09-16 13:36:08 -03:00

autoCombo

fix(providers): remove the chipotle/pepper provider (#13131) (#13913)

2026-09-17 13:22:09 -03:00

browserBackedChat

feat(providers): refresh Z.ai Web models and browser transport (#12524)

2026-09-03 07:49:42 -03:00

codexAccount

fix(codex): lift nested child cooldowns on parent clear and on snapshot headroom (#12951)

2026-09-10 08:15:57 -03:00

combo

fix(combo): retry pre-content streaming failures (#13630)

2026-09-17 02:31:36 -03:00

compression

fix(compression): skip RTK dedup and truncation for non-shell tool results (#13521)

2026-09-17 12:38:16 -03:00

promptql

fix(sse): replace spoofable .includes() PromptQL issuer check with hostname comparison (#8029) (#8042)

2026-07-21 21:42:36 -03:00

rateLimitManager

fix(resilience): parse RFC 3339 and fractional-second rate-limit resets (#13321)

2026-09-16 21:02:04 -03:00

routing

fix(quota): bound routing caches with shared boundedMap factory (#13280)

2026-09-15 13:08:21 -03:00

sessionPool

fix(providers): retire common ChatGPT Web provider (#11754)

2026-08-28 06:52:46 -03:00

tokenRefresh

fix(oauth): bind Google refresh to the client that issued the token (#12106)

2026-08-30 11:10:44 -03:00

usage

feat(usage): show separate Fable weekly limits (#13266)

2026-09-17 16:23:46 -03:00

accountFallback.ts

fix(resilience): lock the exact model, not the quota family, on 5xx model-lockout failures (#12957)

2026-09-17 16:25:32 -03:00

accountSelector.ts

Potential fix for code scanning alert no. 54: Insecure randomness

2026-03-05 08:41:44 -03:00

accountSemaphore.ts

fix(codex): fail fast and release per-account Responses WS leases (#12911)

2026-09-17 10:56:01 -03:00

adaptiveEffort.ts

feat: adaptive reasoning effort (auto) — gateway-resolved, per-turn pinned, all harnesses (#13448)

2026-09-16 15:18:36 -03:00

adobeFireflyAccount.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyArp.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyBrowserLogin.ts

fix(open-sse): write Adobe Firefly session tokens and cookie jars with 0700/0600 perms (#12572) (#13242)

2026-09-11 22:05:11 -03:00

adobeFireflyCatalog.ts

chore(quality): re-freeze the ESLint suppressions on release/v3.8.51 from a clean-room run (#11955)

2026-08-28 22:07:40 -03:00

adobeFireflyClient.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyCredentials.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyIms.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyModels.ts

fix(adobe-firefly): retain Topaz catalog models

2026-08-09 00:21:52 -03:00

adobeFireflyModelSnapshot.ts

fix(adobe-firefly): sync models and media capabilities

2026-08-03 17:22:08 +03:00

adobeFireflyPayload.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyPoll.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyReferences.ts

fix: clear release unit and quality regressions

2026-08-09 17:54:23 -03:00

adobeFireflyResponses.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflySecurity.ts

fix(release): clear remaining Adobe and typecheck gates

2026-08-09 17:02:47 -03:00

adobeFireflySession.ts

fix(adobe-firefly): do not spawn Chrome for CDP warm under test runners (#13289)

2026-09-17 02:32:10 -03:00

adobeFireflyUpload.ts

refactor(quality): split adobeFireflyClient.ts (2,952 lines) into nine leaf modules (#11728)

2026-08-28 09:30:57 -03:00

adobeFireflyUpscale.ts

fix(release): clear remaining Adobe and typecheck gates

2026-08-09 17:02:47 -03:00

agentrouterQuotaFetcher.ts

feat(sse): quota tracking for AgentRouter, v0 (Vercel), FreeModel (#6850, #6845, #7075) (#7653)

2026-07-18 03:17:09 -03:00

AGENTS.md

Fix routed target request parameters (#7323)

2026-07-18 21:18:34 -03:00

aihordeImageCatalog.ts

fix(config): keep the SQLite driver out of the client bundle (#10692) (#10695)

2026-08-18 19:23:46 -03:00

alibabaFreeTier.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

alibabaFreeTierAllowlist.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

alibabaFreeTierDiscovery.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

alibabaFreeTierQuotaClassify.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

alibabaFreeTierQuotaFetcher.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

alibabaFreeTierQuotaTypes.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

antigravity429Engine.ts

fix: treat zero-reset Antigravity 429s as transient (#8626)

2026-08-06 06:05:58 -03:00

antigravityClientProfile.ts

refactor(antigravity): align official clients and callable catalog (#8013)

2026-07-22 20:41:19 -03:00

antigravityCredits.ts

refactor(antigravity): align official clients and callable catalog (#8013)

2026-07-22 20:41:19 -03:00

antigravityFamilyCooldown.ts

fix(quota): keep Antigravity Gemini usable when Claude weekly is empty (#12566)

2026-09-03 12:40:21 -03:00

antigravityHeaders.ts

fix(antigravity): send complete loadCodeAssist metadata (ideType/platform/pluginType as numeric enums) (#11969)

2026-08-30 09:09:51 -03:00

antigravityHeaderScrub.ts

Release v3.8.40

2026-06-29 08:40:06 -03:00

antigravityIdentity.ts

fix(antigravity): derive stable session ID from credentials for prompt-cache reuse (#13305) (#13424)

2026-09-14 23:56:35 -03:00

antigravityProjectBootstrap.ts

fix(providers): poll onboardUser LRO before classifying BYOP (#11519)

2026-08-26 09:21:41 -03:00

antigravityProjectPersist.ts

fix(responses-continuation): recover a real id/output for passthrough and translate-mode replies (#11434)

2026-08-24 19:57:12 -03:00

antigravityProjectPersistence.ts

fix(responses-continuation): recover a real id/output for passthrough and translate-mode replies (#11434)

2026-08-24 19:57:12 -03:00

antigravityQuotaFamily.ts

fix(combo): keep Antigravity Gemini usable when Claude weekly is empty (#12637)

2026-09-03 23:47:28 -03:00

antigravityVersion.ts

refactor(antigravity): align official clients and callable catalog (#8013)

2026-07-22 20:41:19 -03:00

apiKeyRotator.ts

fix(sse): scope credential health to exact key (#11628)

2026-08-26 09:21:52 -03:00

autoRefreshDaemon.ts

fix(combo): guarantee combo loops terminate with an actionable error instead of hanging silently (#10463)

2026-08-20 23:07:25 -03:00

backgroundTaskDetector.ts

chore(lifecycle): gate DEFAULT_DEGRADATION_MAP against retired ids (#12535)

2026-09-11 17:42:16 -03:00

bailianQuotaFetcher.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

batchProcessor.ts

fix(db): cap per-request batch sweep and guard shared files (#13680, #13681) (#13805)

2026-09-16 06:10:43 -03:00

bedrock.ts

Release v3.8.4 (#2678)

2026-05-26 23:51:47 -03:00

bottleneckPatch.ts

fix(ci): clear base-reds on release/v3.8.50 (round 3) (#10213)

2026-08-13 00:02:25 -03:00

browserBackedChat.ts

feat(providers): refresh Z.ai Web models and browser transport (#12524)

2026-09-03 07:49:42 -03:00

browserPool.ts

perf(stream): compile hot-path regexes once, bound token caches, fix quadratic buffering (#12179)

2026-09-03 08:29:14 -03:00

ccBridgeTransforms.ts

fix(providers): add CLAUDE_CODE_CLIENT_VERSION and GITHUB_COPILOT_CLI_VERSION env overrides (#12632)

2026-09-03 23:43:38 -03:00

ccOpenAiMediaBlocks.ts

fix(sse): CC bridge loses OpenAI-format image input (OpenCode/Kilo/Cline → AgentRouter) (#7888)

2026-07-20 15:56:40 -03:00

ccWireImageBuiltins.ts

Release v3.8.45 (#6202)

2026-07-06 02:25:17 -03:00

chatgptWebCodexAdmin.ts

Add native ChatGPT Web provider for Codex clients (#8949)

2026-08-11 09:53:39 -03:00

claudeAdaptiveThinking.ts

fix(types): preserve Claude thinking body contracts (#9852)

2026-08-09 09:51:56 -03:00

claudeCodeCCH.ts

fix(open-sse): serialize xxhash startup in cch hashing

2026-04-13 14:43:32 -03:00

claudeCodeCompatible.ts

fix(providers): add CLAUDE_CODE_CLIENT_VERSION and GITHUB_COPILOT_CLI_VERSION env overrides (#12632)

2026-09-03 23:43:38 -03:00

claudeCodeCompatibleBeta.ts

Release v3.8.26 (#3875)

2026-06-16 01:00:40 -03:00

claudeCodeCompatibleThinkingDisplay.ts

Release v3.8.38 (#5078)

2026-06-27 09:07:12 -03:00

claudeCodeConstraints.ts

fix(sse): advance Claude cache breakpoints on growing tails (#10684)

2026-08-20 13:05:46 -03:00

claudeCodeExtraRemap.ts

Release v3.6.9 (#1404)

2026-04-19 19:50:30 -03:00

claudeCodeFingerprint.ts

feat(claude-code): PR #1188 — Native parity with CCH signing, tool remapping, and API constraints

2026-04-12 22:41:12 -03:00

claudeCodeObfuscation.ts

fix(claude): preserve signed thinking turns during obfuscation (#8629)

2026-08-06 05:24:13 -03:00

claudeCodeToolRemapper.ts

fix(translator): preserve exact tool name casing for Claude tool_use payloads (#11487) (#11520)

2026-08-25 19:25:18 -03:00

claudeHaikuConstraints.ts

Release v3.8.36 (#4854)

2026-06-25 13:17:40 -03:00

claudeTlsClient.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

claudeTurnstileSolver.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

claudeUsageCooldown.ts

Release v3.8.36 (#4854)

2026-06-25 13:17:40 -03:00

claudeWebAutoRefresh.ts

feat(claude-web): implement session-based executor with auto-refresh (#2283)

2026-05-15 12:51:07 -03:00

clinepassModels.ts

feat(cline): align ClinePass catalog and request protocol (#7914)

2026-07-21 12:19:01 -03:00

cloudCodeHeaders.ts

fix(cli): add Gemini CLI fingerprint (#1866)

2026-05-02 00:48:54 -03:00

cloudCodeThinking.ts

perf(proxy): implement non-blocking async proxy log batching and performance optimizations (A, B, C, D) (#11182)

2026-08-23 01:07:32 -03:00

codeAssistSubscription.ts

Release v3.8.2 (#2503)

2026-05-23 01:46:59 -03:00

codexQuotaFetcher.ts

feat(codex): sync v178 identity mechanisms — turn-state relay, persisted seeds, identity faces (#10716)

2026-08-20 06:28:45 -03:00

codexUsageQuotas.ts

fix(quota): align AUTH window labels with usage API and clarify 503 (#12884)

2026-09-10 10:47:11 -03:00

codexVerbosity.ts

Release v3.8.44 (#5925)

2026-07-04 13:00:30 -03:00

combo.ts

fix(combo): retry pre-content streaming failures (#13630)

2026-09-17 02:31:36 -03:00

comboAgentMiddleware.ts

feat(sse): server-side template expansion for combo system prompts (#5501) (#9414)

2026-08-11 04:35:20 -03:00

comboConfig.ts

feat(combo): quota-weighted routing — skip empty accounts, draw by leftover (#12789)

2026-09-07 09:05:29 -03:00

comboContextLimit.ts

fix(combo): honor an operator-set context_length at request time (#12090)

2026-08-30 11:44:18 -03:00

comboMetrics.ts

Release v3.8.34 (#4614)

2026-06-23 03:08:29 -03:00

conolAuth.ts

feat(providers): add Conol (conol.ai) web session provider (#8974)

2026-08-11 09:08:23 -03:00

conolBrowserLogin.ts

feat(providers): add Conol (conol.ai) web session provider (#8974)

2026-08-11 09:08:23 -03:00

conolModels.ts

fix(providers): eliminate Gemini 3.5 Flash

2026-08-23 18:13:11 -03:00

conolSessionModel.ts

feat(providers): add Conol (conol.ai) web session provider (#8974)

2026-08-11 09:08:23 -03:00

conolUsage.ts

feat(providers): add Conol (conol.ai) web session provider (#8974)

2026-08-11 09:08:23 -03:00

contextHandoff.ts

fix(combo): universal-handoff fixes — bare-fallback note, same-request scoping, silent-failure logging (#12338)

2026-09-03 13:02:24 -03:00

contextManager.ts

fix(db): bound health scans and isolate native diagnostics (#13717)

2026-09-16 12:59:45 -03:00

conversationTracker.ts

feat(sse): allow disabling conversation tracking (#13150)

2026-09-17 16:24:21 -03:00

conversationTurnContent.ts

fix(conversations): resolve turn content OmniRoute never sends back to the client (#12447)

2026-09-03 13:00:44 -03:00

credentialGate.ts

Release v3.8.3 (#2617)

2026-05-24 18:05:58 -03:00

crofUsageFetcher.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

cursorApiKeyAuth.ts

fix(security): clear new CodeQL code-scanning alerts (round 2) (#10888)

2026-08-21 08:02:16 -03:00

cursorSessionManager.ts

fix(cursor): preserve tool context across multi-turn conversations when client lacks conversation_id (#9029)

2026-08-07 11:23:27 -03:00

dailyQuotaReset.ts

fix(sse): bound daily quota cooldowns around daylight-saving transitions (#13671)

2026-09-15 15:07:42 -03:00

dashscopeTextModels.ts

feat(alibaba): free-tier routing with live quota sync (#8893)

2026-08-11 04:25:23 -03:00

deepseekQuotaFetcher.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

defaultReasoningEffort.ts

fix(models): enforce a synced model's real context window and default effort immediately (#10957)

2026-08-21 08:25:14 -03:00

deviceTracker.ts

Release v3.8.44 (#5925)

2026-07-04 13:00:30 -03:00

emergencyFallback.ts

Release v3.8.24 (#3747)

2026-06-13 17:27:40 -03:00

errorClassifier.ts

fix(sse): classify a 2xx body as a disguised upstream failure (#13461) (#13910)

2026-09-17 10:45:25 -03:00

evalRouting.ts

Release v3.8.4 (#2678)

2026-05-26 23:51:47 -03:00

firecrawlQuotaFetcher.ts

Merge remote-tracking branch 'origin/release/v3.8.50' into fix/release-v3.8.50-post-sweep-base-red

2026-08-09 15:21:37 -03:00

freeModelQuotaFetcher.ts

feat(sse): quota tracking for AgentRouter, v0 (Vercel), FreeModel (#6850, #6845, #7075) (#7653)

2026-07-18 03:17:09 -03:00

freeWebSearch.ts

Release v3.8.40

2026-06-29 08:40:06 -03:00

fusion.ts

feat: combo-lane awareness + activation UX + MCP visibility (Wave 2 of #9654) (#10039)

2026-08-18 11:31:46 -03:00

geminiRateLimitTracker.ts

fix(sse): Gemini TPM/RPD quota classification + combo cooldown-wait resilience (#8213)

2026-07-23 05:17:35 -03:00

geminiThoughtSignatureStore.ts

perf(ram): reduce server memory footprint (#2903)

2026-05-29 13:21:36 -03:00

genericQuotaFetcher.ts

fix(quota): bound routing caches with shared boundedMap factory (#13280)

2026-09-15 13:08:21 -03:00

gigachatAuth.ts

perf(stream): compile hot-path regexes once, bound token caches, fix quadratic buffering (#12179)

2026-09-03 08:29:14 -03:00

githubCopilotModels.ts

feat(providers): filter GitHub combo members against live catalog (#12473)

2026-09-03 23:39:43 -03:00

gpt5SamplingGuard.ts

fix(openai): strip reasoning_effort when GPT-5.x models carry function tools (#7101)

2026-07-17 10:43:37 -03:00

grokClearance.ts

feat: classify grok-web Cloudflare anti-bot blocks + gated browser-backed cf_clearance path (#8019) (#8241)

2026-07-23 01:33:32 -03:00

grokCliQuotaFetcher.ts

feat: add live gRPC-web quota fetcher for grok-cli (#6844) (#7714)

2026-07-20 15:57:29 -03:00

grokCliQuotaFrame.ts

feat: add live gRPC-web quota fetcher for grok-cli (#6844) (#7714)

2026-07-20 15:57:29 -03:00

grokQuotaFetcher.ts

feat(providers): weekly quota for xAI OAuth (Grok) (#8471)

2026-07-26 12:11:02 -03:00

grokResetCredits.ts

feat(grok-cli): show and redeem banked reset credits on Provider Limits (#12805)

2026-09-07 08:58:12 -03:00

grokResetCreditsFrame.ts

feat(grok-cli): show and redeem banked reset credits on Provider Limits (#12805)

2026-09-07 08:58:12 -03:00

grokTlsClient.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

hfModelSuggestions.ts

Release v3.8.44 (#5925)

2026-07-04 13:00:30 -03:00

httpBackedChatFingerprint.ts

fix(claude-web): unify Turnstile/executor/fast-path User-Agents behind one fingerprint (#7548) (#7711)

2026-07-19 02:31:26 -03:00

hyperagentModels.ts

fix(hyperagent): default 1M context for fable/opus/sonnet (#8496)

2026-07-26 03:53:06 -03:00

imageCombo.ts

fix(images): fall through combo edit targets on /v1/images/edits (#12653)

2026-09-11 17:41:33 -03:00

inAppLoginService.ts

chore(release): merge release/v3.8.50 tip into release/v3.8.51 — sync-back step 1/2

2026-08-28 14:07:39 -03:00

intentClassifier.ts

Release v3.8.2 (#2503)

2026-05-23 01:46:59 -03:00

ipFilter.ts

fix(security): empty whitelist allows through to prevent admin lockout (#13534)

2026-09-14 23:55:21 -03:00

keyGroupAuth.ts

fix(barrel): migrate open-sse, src/shared, src/sse, src/models, src/domain off the @/lib/localDb barrel import (#11795 Phase 4) (#12053)

2026-08-30 02:31:09 -03:00

kiroExternalIdp.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

kiroModels.ts

fix(kiro): read usage from the frames Kiro actually sends (#9035)

2026-08-05 21:42:52 -03:00

kiroRegion.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

learnedReasoningEffortCaps.ts

fix(providers): unify learned and declared reasoning-effort clamps on nearest-tier semantics (#11295) (#11305)

2026-08-23 22:17:12 -03:00

learnedThinkingCaps.ts

Train 1D: merge via --admin on .113 validation

2026-07-27 11:30:33 -03:00

lmarenaTlsClient.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

manifestAdapter.ts

fix(routing): table pricing with catalog fallback for off-table models, pooled latency bootstrap, fresh tier cache (#12792)

2026-09-10 09:20:18 -03:00

maxaiModels.ts

feat(providers): add MaxAI — signed OpenAI-compatible provider (chat, tools, vision, image-gen, doc-RAG) (#11461)

2026-09-02 01:55:42 -03:00

mimoThinking.ts

Release v3.8.29 (#4126)

2026-06-19 06:49:01 -03:00

model.ts

fix(providers): stop zed-hosted claude-haiku-4-5 thinking from inflating max_tokens (#13364) (#13780)

2026-09-16 06:15:00 -03:00

modelCapabilities.ts

feat: v3.6.4 — Combo Builder v2, Composite Tiers, P2C Credentials, Observability Layer

2026-04-12 10:34:10 -03:00

modelDeprecation.ts

feat(providers): Agnes 3.0 Flash catalog, CN host, live /v1/models (#13120)

2026-09-11 19:28:10 -03:00

modelEndpointPolicy.ts

fix: enforce OpenAI model lifecycle without silent reroutes (#8627)

2026-08-12 09:04:41 -03:00

modelFamilyFallback.ts

feat(providers): refresh Fable, Cursor, and Devin catalogs (#12367)

2026-09-03 07:50:41 -03:00

modelLifecycle.ts

fix(autoCombo,sse): drop vendor-retired ids from the auto-combo pool and skip leftover arena_elo (#11625) (#11626)

2026-08-26 09:25:25 -03:00

modelscopePolicy.ts

Release v3.8.6 (#2804)

2026-05-29 12:44:29 -03:00

modelStrip.ts

feat(release): implement Hermes CLI config and message content stripping (#1475, #1387)

2026-04-22 01:39:49 -03:00

moonshotQuotaFetcher.ts

fix(dashboard): moonshot voucher/cash leftover follows bucket balance (#12733)

2026-09-07 08:57:02 -03:00

newApiAggregatorQuotaFetcher.ts

feat(sse): New-API/One-API/Sub2API aggregator balance detection (#9415) (#9539)

2026-08-06 10:40:20 -03:00

notionStreamParser.ts

fix(notion-web): use Chrome TLS impersonation for runInferenceTranscript (#8159)

2026-07-23 12:17:58 -03:00

notionThreadSessions.ts

Stabilize Notion web sessions and JSON output (#8751)

2026-07-27 17:25:43 -03:00

notionTlsClient.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

notionTranscriptBuilder.ts

fix(notion-web): reuse threadId across OpenAI multi-turn (no new chat each request) (#7900)

2026-07-22 06:59:24 -03:00

notionWebFallbackModels.ts

feat(providers): refresh curated model catalogs and retire Imagen 4 (#10537)

2026-08-18 12:27:46 -03:00

notionWebModels.ts

fix(notion-web): add browser fingerprint headers to reduce Cloudflare challenges (#7864)

2026-07-20 15:57:15 -03:00

oauthSessionOccupancy.ts

fix(routing): account for active OAuth sessions (#8940)

2026-08-11 09:51:35 -03:00

obscura.ts

feat(browser): adopt Obscura as primary headless browser engine with Chromium fallback (#12286)

2026-09-03 08:24:32 -03:00

opencodeOllamaUsage.ts

fix(sse): parse Ollama Cloud's current usage markup (#12749) (#13260)

2026-09-11 22:06:29 -03:00

opencodeQuotaFetcher.ts

feat(quota): use official OpenCode Go usage API (#12124)

2026-08-30 11:11:46 -03:00

opencodeReasoningSanitizer.ts

feat(sse): support native max reasoning effort and per-model clamping (#11875)

2026-09-02 00:12:13 -03:00

openrouterFreeWindow.ts

feat: OpenRouter quota tracking (key/credits + free-window counter) (#6842) (#7651)

2026-07-18 03:11:01 -03:00

openrouterQuotaFetcher.ts

feat(usage): devin-cli agentic quota + openrouter credits in Provider Limits (#12256)

2026-09-02 00:02:15 -03:00

payloadRules.ts

fix(barrel): migrate open-sse, src/shared, src/sse, src/models, src/domain off the @/lib/localDb barrel import (#11795 Phase 4) (#12053)

2026-08-30 02:31:09 -03:00

perplexityTlsClient.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

pipeline.ts

fix(combo): exclude hidden leaves from catalog and dispatch (#8878)

2026-08-06 00:08:52 -03:00

promptqlModels.ts

fix(providers): eliminate Gemini 3.5 Flash

2026-08-23 18:13:11 -03:00

provider.ts

fix(translator): harden Claude format detection for model validation (#9253)

2026-08-04 10:06:31 -03:00

providerAlias.ts

fix(build): stop the client bundle from reaching server-only modules, and make the guard find them (#13436)

2026-09-15 12:52:41 -03:00

providerCooldownTracker.ts

feat(sse): wire the PROVIDER_PROFILES window gate into the global provider cooldown (#12247)

2026-09-01 01:57:17 -03:00

providerCostData.ts

fix(routing): table pricing with catalog fallback for off-table models, pooled latency bootstrap, fresh tier cache (#12792)

2026-09-10 09:20:18 -03:00

providerDefaultRateLimit.ts

fix(resilience): stop unbounded queue that hangs 6min until Aborted (#12715)

2026-09-10 10:49:34 -03:00

providerRequestDefaults.ts

Release v3.8.0 (#2073)

2026-05-10 00:55:06 -03:00

proxyAutoSelector.ts

Release v3.8.14 (#3340)

2026-06-07 07:20:02 -03:00

qoderCli.ts

fix(windows): request shell when spawning bare qoder binary name on Windows (fixes #8590) (#8633)

2026-07-27 19:08:12 -03:00

qoderCliResolve.ts

fix(windows): request shell when spawning bare qoder binary name on Windows (fixes #8590) (#8633)

2026-07-27 19:08:12 -03:00

quotaFetchThrottle.ts

Release v3.8.44 (#5925)

2026-07-04 13:00:30 -03:00

quotaMonitor.ts

feat(sse): New-API/One-API/Sub2API aggregator balance detection (#9415) (#9539)

2026-08-06 10:40:20 -03:00

quotaPreflight.ts

fix(claude): extra-usage switch does not skip 5h preflight (#12803)

2026-09-07 08:59:39 -03:00

quotaResetParsing.ts

fix(responses-continuation): recover a real id/output for passthrough and translate-mode replies (#11434)

2026-08-24 19:57:12 -03:00

quotaTextCooldowns.ts

fix(search): treat HTTP 432 and plan limit errors as transient cooldown (#12139)

2026-08-30 19:39:36 -03:00

quotaTrackersBatch.ts

feat(providers): live monthly credit quota for Firecrawl (#8759)

2026-07-27 17:25:47 -03:00

quotaWindowLabel.ts

fix(quota): align AUTH window labels with usage API and clarify 503 (#12884)

2026-09-10 10:47:11 -03:00

qwenThinking.ts

feat(core): add db health checks and provider interoperability

2026-04-13 13:11:30 -03:00

qwenTokenPlanQuotaFetcher.ts

fix(sse): surface Qwen/Alibaba personal Token Plan quota in dashboard and preflight (#10290)

2026-08-14 15:40:33 -03:00

rateLimitManager.ts

fix(quota): bound routing caches with shared boundedMap factory (#13280)

2026-09-15 13:08:21 -03:00

rateLimitSemaphore.ts

fix(sse): keep a round-robin target's concurrency limit through a cooldown (#13320)

2026-09-16 21:01:48 -03:00

reasoningCache.ts

fix(reasoning): expand deepseek-v4 regex to match v4.1+ and add command-code provider (#13430) (#13528)

2026-09-15 00:06:58 -03:00

reasoningInputPolicy.ts

fix(translator): strip plaintext reasoning content for opaque responses backends (#12128) (#12171)

2026-08-31 14:10:26 -03:00

reasoningTokenBuffer.ts

feat(reasoning): opt-in min output budget floor for thinking models (#12742)

2026-09-16 13:38:23 -03:00

refreshSerializer.ts

fix(types): narrow refresh token rotation inputs (#10257)

2026-08-14 00:57:58 -03:00

requestDedup.ts

docs(dedup): correct the CodeQL claim on the tenant-namespace comment (#11712)

2026-08-28 01:57:27 -03:00

responseModelEcho.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

responsesInputSanitizer.ts

perf(stream): compile hot-path regexes once, bound token caches, fix quadratic buffering (#12179)

2026-09-03 08:29:14 -03:00

responsesItemId.ts

fix(sse): default summary + strip malformed id on kept Responses input items (#11110)

2026-08-22 14:39:47 -03:00

retryAfterJson.ts

fix(antigravity): cap unverified body retry hints (#11823)

2026-08-28 11:14:24 -03:00

roleNormalizer.ts

fix(ddgw): normaliza system no upstream da release v3.8.51 (#11692)

2026-08-26 17:41:28 -03:00

rollingRpmGate.ts

fix(resilience): enforce RPM with rolling leases (#9604)

2026-08-06 10:41:04 -03:00

rotationConfig.ts

Release v3.8.47 (#6569)

2026-07-13 09:12:40 -03:00

searchCache.ts

perf(ram): reduce server memory footprint (#2903)

2026-05-29 13:21:36 -03:00

sessionManager.ts

Release v3.8.10 (#3140)

2026-06-04 20:05:38 -03:00

signatureCache.ts

perf(ram): reduce server memory footprint (#2903)

2026-05-29 13:21:36 -03:00

slidingWindowLimiter.ts

fix(resilience): enforce RPM with rolling leases (#9604)

2026-08-06 10:41:04 -03:00

specificityDetector.ts

Release v3.8.0 (#2073)

2026-05-10 00:55:06 -03:00

specificityRules.ts

Release v3.8.0 (#2073)

2026-05-10 00:55:06 -03:00

specificityTypes.ts

fix(types): preserve request rule contracts (#9135)

2026-08-05 22:31:10 -03:00

speechCombo.ts

chore(audio): retire EdgeTTS pending provenance review (#11711)

2026-08-28 04:45:17 -03:00

streamRecovery.ts

fix(stream-recovery): log recovery traces with continuation attempt (#13650)

2026-09-15 16:19:10 -03:00

systemPrompt.ts

fix(sse): inject global system prompt once, post-translation, across all target shapes (#12904)

2026-09-17 02:30:35 -03:00

systemTransforms.ts

fix(sse): do not ZWJ-obfuscate the substring hermes in user text (#10488)

2026-08-17 07:04:48 -03:00

targetRequestSanitizer.ts

fix(sse): guard empty tool_calls[] and strip tool_choice without tools (#12901)

2026-09-17 02:31:19 -03:00

taskAwareRouter.ts

feat(auto-combo): declare observed reliability as a scoring factor (#12317)

2026-09-01 15:36:39 -03:00

taskAwareRouting.ts

fix(responses-continuation): recover a real id/output for passthrough and translate-mode replies (#11434)

2026-08-24 19:57:12 -03:00

thinkingBudget.ts

fix(db): bound health scans and isolate native diagnostics (#13717)

2026-09-16 12:59:45 -03:00

throughputWatchdog.ts

feat(resilience): abort persistently slow upstream streams (#9709) (#10012)

2026-08-10 18:16:13 -03:00

tierConfig.ts

fix(providers): reclassify Cerebras as a one-time $5 signup credit (#12591)

2026-09-03 21:30:32 -03:00

tierDefaults.json

fix(providers): reclassify Cerebras as a one-time $5 signup credit (#12591)

2026-09-03 21:30:32 -03:00

tierResolver.ts

fix(routing): table pricing with catalog fallback for off-table models, pooled latency bootstrap, fresh tier cache (#12792)

2026-09-10 09:20:18 -03:00

tierTypes.ts

Release v3.8.0 (#2073)

2026-05-10 00:55:06 -03:00

tlsClientBase.ts

fix(providers): migrate web cookie TLS transport to wreq-js (#12429)

2026-09-02 10:41:00 -03:00

tlsClientProxy.ts

Release v3.8.24 (#3747)

2026-06-13 17:27:40 -03:00

tokenExtractionConfig.ts

fix(sync): repair the two regressions the v3.8.50 sync-back left on release/v3.8.51

2026-08-28 16:41:42 -03:00

tokenLimitCounter.ts

fix(barrel): migrate open-sse, src/shared, src/sse, src/models, src/domain off the @/lib/localDb barrel import (#11795 Phase 4) (#12053)

2026-08-30 02:31:09 -03:00

tokenRefresh.ts

fix(oauth): bind Google refresh to the client that issued the token (#12106)

2026-08-30 11:10:44 -03:00

toolLatencyTracker.ts

Release v3.8.10 (#3140)

2026-06-04 20:05:38 -03:00

toolLimitDetector.ts

Release v3.8.45 (#6202)

2026-07-06 02:25:17 -03:00

toolSchemaSanitizer.ts

fix(kimi): normalize MFJS tool schemas (#10079)

2026-08-13 07:52:41 -03:00

transientBackendRetry.ts

feat(sse): retry transient 5xx backend errors with jitter (global-fallback call site) (#12695) (#13143)

2026-09-17 10:43:54 -03:00

usage.ts

feat(quota): Moonshot Open Platform balance and TPD lock for custom nodes (#12590)

2026-09-03 12:47:50 -03:00

v0QuotaFetcher.ts

feat(sse): quota tracking for AgentRouter, v0 (Vercel), FreeModel (#6850, #6845, #7075) (#7653)

2026-07-18 03:17:09 -03:00

videoCombo.ts

fix(providers): resolve combo names on /v1/audio/speech and /v1/videos/generations (#10471)

2026-08-18 10:50:31 -03:00

volcengineConsoleAutoLogin.ts

fix(sse): exact-domain cookie match and origin-equality URL assertions (#11452)

2026-08-24 23:46:45 -03:00

volumeDetector.ts

fix(barrel): migrate open-sse, src/shared, src/sse, src/models, src/domain off the @/lib/localDb barrel import (#11795 Phase 4) (#12053)

2026-08-30 02:31:09 -03:00

wafRateLimit.ts

fix(agentrouter): retry on 400 content-blocked + burst guard (#9323)

2026-08-03 18:22:14 -03:00

webFetchInterception.ts

fix(types): align web fallback contracts (#8819)

2026-07-27 23:57:23 -03:00

webSearchFallback.ts

fix(providers): bypass web-search fallback for Antigravity target (#13447) (#13765)

2026-09-16 13:29:05 -03:00

webSearchRouting.ts

Release v3.8.33 (#4515)

2026-06-22 03:17:02 -03:00

webSessionPoolHealth.ts

feat: add web-session pool observability (MCP tool + health-matrix) (#3395)

2026-06-08 00:48:01 -03:00

wildcardRouter.ts

fix(build): unblock release build and settings state updates

2026-04-17 23:21:02 -03:00

workflowFSM.ts

chore(release): v3.4.8 — security compliance and fix algorithms

2026-04-03 00:15:35 -03:00

xaiMessageCap.ts

fix(xai): cap chat history at xAI 800-message limit (#10601)

2026-08-18 10:52:52 -03:00

xaiOauthQuotaFetcher.ts

feat(providers): weekly quota for xAI OAuth (Grok) (#8471)

2026-07-26 12:11:02 -03:00

zaiWebCredentials.ts

fix(build): repair broken production build, red lint gate and SWR crash (#10198)

2026-08-12 15:58:40 -03:00