fix(routing): bare model ids route to codex first; validate synced candidates by diegosouzapw · Pull Request #9275 · diegosouzapw/OmniRoute

diegosouzapw · 2026-08-03T05:48:43Z

Two bare-model-routing bugs surfaced in the field when an OmniRoute deployment had a codex subscription whose cookie quota was exhausted (retry-after 429047s / ~5 days) AND an active kiro connection whose upstream sync briefly advertised 'claude-opus-5' before kiro vendored it into the static registry.

Bug 1: Bare 'gpt-5.6-sol' routes to codex even when codex is exhausted

Bare 'gpt-5.6-sol' (and friends) routed to the codex provider via the 'codex-only installations' branch of resolveModelByProviderInference — even when the user had explicitly configured 'agentrouter' as their model_provider in the codex CLI config. With codex in cooldown (retry-after: 429047s ≈ 5 days), every bare request 429'd.

Fix: extend CODEX_NATIVE_UNPREFIXED_MODELS to include the full gpt-5.6-sol tier set + gpt-5.5 + the related codex-native ids. The Codex CLI default is now actually honored; users can still prefix 'agentrouter/gpt-5.6-sol' to opt into a specific provider (existing regression-tested).

Bug 2: Bare 'claude-opus-5' silently routes to kiro (synced-catalog phantom)

Bare 'claude-opus-5' silently routed to 'kiro' when kiro's synced /v1/models catalog had that id (transient upstream quirk). kiro's static registry never cataloged claude-opus-5, so the upstream call 404'd with 'No active credentials for provider: kiro'.

Fix: validate activeSyncedProviders against MODEL_TO_PROVIDERS before merging them into the candidate list. Auto-discovery still wins when the model id has no static entry (brand-new models from upstream keep working).

Bonus: error message now includes candidate-prefix hint

When handleNoCredentials returns a 404 'No active credentials for provider: X', surface the top-3 candidate aliases (e.g. 'anthropic/claude-opus-5, claude/claude-opus-5, agentrouter/claude-opus-5') so the operator can pick a working prefix instead of staring at a wall.

Files changed

open-sse/services/model.ts — CODEX_NATIVE_UNPREFIXED_MODELS expanded + synced validation
src/sse/handlers/chatHelpers.ts — handleNoCredentials accepts candidateAliases
src/sse/handlers/chat.ts — passes resolved.candidateAliases to handleNoCredentials
tests/unit/fix-bare-model-precedence.test.ts (7 tests)
tests/unit/fix-synced-model-validation.test.ts (3 tests)
tests/unit/fix-error-message-candidates.test.ts (3 tests)
tests/unit/fix-bare-routing-fallback.test.ts (7 tests)

Validation

npm run typecheck:core → ✅
All 4 new test files: 20 tests, all pass
model-resolver.test.ts: 25 tests, all pass (regression)
bare-model-autopick.test.ts: 13 tests, all pass (regression)
agentrouter-chatcore-protocols.test.ts + agentrouter-models-discovery-7016.test.ts: all pass
chat-cooldown-aware-retry.test.ts: all pass
npm run lint: 2 pre-existing errors in @/lib/localDb imports (untouched by this PR — already in eslint-suppressions.json)

…ndidates Two bare-model-routing bugs surfaced in the field when an OmniRoute deployment had a codex subscription whose cookie quota was exhausted (retry-after 429047s / ~5 days) AND an active kiro connection whose upstream sync briefly advertised 'claude-opus-5' before kiro vendored it into the static registry. 1. Bare 'gpt-5.6-sol' (and friends) routed to the codex provider even when the user had explicitly configured 'agentrouter' as their provider (via model_provider in codex CLI). With codex in cooldown, every bare request 429'd. Fix: extend CODEX_NATIVE_UNPREFIXED_MODELS to include the full gpt-5.6-sol tier set + gpt-5.5 + the related codex-native ids. The Codex CLI default is now actually honored; users can still prefix 'agentrouter/gpt-5.6-sol' to opt into a specific provider. 2. Bare 'claude-opus-5' silently routed to 'kiro' when kiro's synced /v1/models catalog had that id (likely from a transient upstream quirk). kiro's static registry never cataloged claude-opus-5, so the upstream call 404'd. Fix: validate activeSyncedProviders against MODEL_TO_PROVIDERS before merging them into the candidate list. Auto-discovery still wins when the model id has no static entry (brand-new models from upstream keep working). Bonus: when handleNoCredentials returns a 404 'No active credentials for provider: X' error, surface the top-3 candidate aliases (e.g. 'anthropic/claude-opus-5, claude/claude-opus-5, agentrouter/claude-opus-5') so the operator can pick a working prefix instead of staring at a wall. Tests (all pass, 25 regression tests preserved): - tests/unit/fix-bare-model-precedence.test.ts (7 tests) - tests/unit/fix-synced-model-validation.test.ts (3 tests) - tests/unit/fix-error-message-candidates.test.ts (3 tests) - tests/unit/fix-bare-routing-fallback.test.ts (7 tests)

…r WAF The agentrouter.org WAF blocks requests containing 'lorem ipsum' in messages[].content. When Claude Code reads test files via the Read tool, the content appears in tool_result blocks which can trigger the filter. Replace 'lorem ipsum dolor sit amet' with 'example content for testing purposes' in compression harness test to avoid false positives.

The agentrouter.org upstream WAF returns 400 content-blocked intermittently when: 1. messages[].content contains a blocked keyword (Lorem ipsum, the phrase 'language model' alone, 'virtual assistant', etc.); or 2. Requests from the same IP/key arrive in a burst, after which the WAF's per-IP suspicion bucket starts blocking content that would normally pass. The bucket relaxes after ~5-10s of idle. Apply three mitigations: 1. Burst guard (open-sse/services/wafRateLimit.ts) Per-bucket (provider+url) gate that enforces a 500ms minimum gap between outbound requests to agentrouter. Configurable via configureWafRateLimit(). Tested in tests/unit/wafRateLimit.test.ts. 2. Reactive retry (BaseExecutor.WAF_RETRY_CONFIG in base.ts) New WAF_RETRY_CONFIG with maxAttempts=2, delayMs=1500, backoffMultiplier=2. When the upstream returns 400 with a body that matches /content[_-]blocked/i, retry the same URL with exponential backoff (1.5s, 3.0s) before falling through to the 429/401/fallback chain. Tested in tests/unit/base-executor-waf-retry.test.ts. 3. Documentation (docs/security/AGENTROUTER_WAF.md) Blocklist of always-blocked and almost-always-blocked patterns, behavior under load, guidance for prompts/tool output, and pointers to the relevant code paths in OmniRoute. These are belt-and-suspenders: the burst guard prevents the WAF from activating on normal traffic, and the reactive retry recovers when it does anyway. Together they should eliminate the intermittent 400 content-blocked that Claude Code sees when running through agentrouter via OmniRoute. Refs #9275 follow-up. Test: 'WAF retry config shape' and 'WAF retry differs from generic' guard the WAF_RETRY_CONFIG contract so future refactors don't accidentally collapse the two retry paths. Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

Two categories of inherited base-branch breakage surfaced when rebasing onto release/v3.8.50's latest tip, both confirmed unrelated to this PR's own diff: - check:file-size: base.ts and chat.ts drifted further past their frozen caps via already-merged commits (7163081 and others) that didn't rebaseline after growing them. Documented and bumped in file-size-baseline.json. - chat-helpers.test.ts: two gpt-5.5 routing assertions predate diegosouzapw#9275 (fix(routing): bare model ids route to codex first), which deliberately made gpt-5.5 route to codex unconditionally, regardless of which other providers are active. Confirmed via diegosouzapw#9275's own commit message and code comments this is intentional, not a regression; verified reproducible on the raw base tip alone, with no changes from this PR involved. Updated both assertions and their names to match the new, intentional default.

Same two categories of inherited base-branch breakage as diegosouzapw#9006 (this PR shares the identical chatHelpers.ts diff), confirmed unrelated to this PR's own diff: - check:file-size: base.ts and chat.ts drifted further past their frozen caps via already-merged commits (7163081 and others) that didn't rebaseline after growing them. - chat-helpers.test.ts: two gpt-5.5 routing assertions predate diegosouzapw#9275 (fix(routing): bare model ids route to codex first), which deliberately made gpt-5.5 route to codex unconditionally. Updated both assertions and their names to match the new, intentional default.

Previous commit fixed the regression by simply omitting gpt-5.5 from CODEX_NATIVE_UNPREFIXED_MODELS, relying on fallthrough to the pre-existing precedence logic below. That's structurally the same failure mode that caused the original bug: a check "just happens" not to apply to a given model id, with correctness depending on nobody re-adding it later without reading a comment 20 lines away from the call site (exactly what happened once already, in diegosouzapw#9275). Adds a second, explicitly-named set, CODEX_NATIVE_MODELS_WITH_OPENAI_PRECEDENCE, and checks it directly at the call site: `CODEX_NATIVE_UNPREFIXED_MODELS.has(modelId) && !CODEX_NATIVE_MODELS_WITH_OPENAI_PRECEDENCE.has(modelId)`. A future engineer reading this exact line sees the carve-out and why it exists, instead of needing to notice gpt-5.5's absence from a list defined well above it. Test updated to assert membership in the new set directly, rather than only asserting absence from the old one.

… guards The two files #9275 added assert that bare gpt-5.5 / gpt-5.6-sol reach codex, but they ran against an empty database — so they also pinned 'codex wins with no codex connection at all', which is the regression #9447 removes. That put them in direct contradiction with plan3-p0 / chat-helpers / codex-gpt55-routing-5887, which assert openai for the very same input: no implementation could satisfy both, which is why the release could not go green. Seeding an active codex connection keeps the contract these files were written to guard (codex beats openai for a Codex-native bare id) while dropping the accidental 'even with no codex configured' half. Cases that need no connection are left as they were: the tier-only ids and codex-auto-review have no alternative provider to preempt, and the explicit-prefix overrides are unaffected.

…codex is active (#9447) * fix(routing): only let Codex-native bare ids preempt a provider when codex is active #9275 widened CODEX_NATIVE_UNPREFIXED_MODELS from a single id to gpt-5.5 plus the gpt-5.6-sol/terra/luna tiers, so bare Codex CLI ids would reach the ChatGPT subscription instead of fanning out to whichever provider won the inference race. The early return it added never consulted the active-provider set, which made the codex-only guard 30 lines below unreachable for every id in the set: if (CODEX_NATIVE_UNPREFIXED_MODELS.has(modelId)) return { provider: "codex", ... } An OpenAI-only install therefore had bare gpt-5.5 routed to codex and failed with 'no active credentials for provider: codex' on a model OpenAI serves, and an install whose codex connection was merely inactive failed identically. This also silently reverted #5887's compatibility boundary. The preference now only PREEMPTS another provider when a codex connection is active. Ids that no other provider catalogs (codex-auto-review) still resolve to codex with no connection at all — there is nothing to preempt and 'no codex credentials' is the honest error. With codex active the preference still beats OpenAI, which is the point of #9275, and an explicit openai/ prefix overrides it either way. Tests: the three assertions that encode the intended #9275 change now expect codex (plus a new one pinning the explicit-prefix override); the rest were already correct and pass again untouched. Adds a regression test for the OpenAI-only case. * docs(changelog): correct fragment id to #9447 * test(routing): seed an active codex connection in the bare-precedence guards The two files #9275 added assert that bare gpt-5.5 / gpt-5.6-sol reach codex, but they ran against an empty database — so they also pinned 'codex wins with no codex connection at all', which is the regression #9447 removes. That put them in direct contradiction with plan3-p0 / chat-helpers / codex-gpt55-routing-5887, which assert openai for the very same input: no implementation could satisfy both, which is why the release could not go green. Seeding an active codex connection keeps the contract these files were written to guard (codex beats openai for a Codex-native bare id) while dropping the accidental 'even with no codex configured' half. Cases that need no connection are left as they were: the tier-only ids and codex-auto-review have no alternative provider to preempt, and the explicit-prefix overrides are unaffected. --------- Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

…9392) #9275 started appending a candidate-alias hint to the zero-active-credentials error and terminated the provider name with a period, so the two sentences read as one message. The two vscode tokenized-route tests still assert the old unterminated string and now fail on every pull request opened against this branch. The Quality Gates workflow only runs on pull_request to release/**, never on push, so the branch itself never re-runs these shards and the drift stayed invisible after the merge. Assert what the handler actually produces. Keeping the comparison exact rather than loosening it to a prefix match is deliberate -- the exact form is what caught the drift. Signed-off-by: Minxi Hou <houminxi@gmail.com>

…gpt-5.5 contract (owner decision) Keeps this PR's surviving value (the diegosouzapw#9275 comment fixes + the documented diegosouzapw#5887×diegosouzapw#9447 interplay at the preemption site + own-growth baseline entry) and drops the behavioral carve-out: the base's diegosouzapw#9447 active-connection bound already delivers the agreed contract (OpenAI while codex is inactive, active codex preempts, openai/ prefix overrides), encoded in codex-gpt55-routing-5887.test.ts. Inherited file-size bumps reverted (release captain domain; diegosouzapw#9355 base-relative mode protects innocent PRs).

Two categories of inherited base-branch breakage surfaced when rebasing onto release/v3.8.50's latest tip, both confirmed unrelated to this PR's own diff: - check:file-size: base.ts and chat.ts drifted further past their frozen caps via already-merged commits (7163081 and others) that didn't rebaseline after growing them. Documented and bumped in file-size-baseline.json. - chat-helpers.test.ts: two gpt-5.5 routing assertions predate diegosouzapw#9275 (fix(routing): bare model ids route to codex first), which deliberately made gpt-5.5 route to codex unconditionally, regardless of which other providers are active. Confirmed via diegosouzapw#9275's own commit message and code comments this is intentional, not a regression; verified reproducible on the raw base tip alone, with no changes from this PR involved. Updated both assertions and their names to match the new, intentional default.

…n every provider (#9006) * fix(executors): route Claude-via-Vertex through native rawPredict with real streaming Claude models on Vertex AI were being sent through the generic OpenAI- compatible partner endpoint, which 404s/errors for Claude on at least some projects. Route them through Vertex's native Anthropic Messages API (publishers/anthropic/.../rawPredict) instead, stripping the body-level model field rawPredict rejects and injecting the required anthropic_version field. rawPredict only ever returns a complete JSON body, never real SSE framing, so streaming requests now get a genuine Anthropic-format SSE stream synthesized from that JSON (message_start/content_block_*/ message_delta/message_stop), which the existing claude-to-openai response translator already knows how to parse. Also fixes two response-format resolution bugs that silently dropped a custom model's DB-stored targetFormat override whenever the model id also existed in the static provider registry (as claude-sonnet-4-6 and claude-opus-4-7 do under vertex): resolveModelOrError had its own ad-hoc resolution that never consulted the override, and even once fixed, executeChatWithBreaker discarded the correctly-resolved format before handleChatCore's own resolution ran a second time. * docs: add changelog fragment for #8909 * refactor(sse): extract shared Claude effort-model predicate * fix(sse): strip Claude effort-suffix ids for any provider serving a real Claude model * fix(sse): keep no-think and CC-discovery catalog variant roots unprefixed * fix(dashboard): re-qualify no-think playground model ids correctly * fix(sse): scope Vertex 404s to a per-model lockout via passthroughModels * docs: add changelog fragment for the Claude catalog/dispatch fix * fix(sse): align regex naming and changelog formatting * fix(sse): clarify effort-variant strip comment and add cross-module drift guard * fix(sse): disambiguate Vertex connection-wide vs per-model 403s * docs: document Vertex 403 disambiguation in changelog fragment * fix(sse): correlate reason and resource within the same ErrorInfo detail * fix(sse): extract Vertex error classifier and rebaseline frozen file sizes * test: register vertex-passthrough-model-lockout in stryker tap.testFiles * fix(sse): reconciles rebase-onto-tip drift for 9006 Two categories of inherited base-branch breakage surfaced when rebasing onto release/v3.8.50's latest tip, both confirmed unrelated to this PR's own diff: - check:file-size: base.ts and chat.ts drifted further past their frozen caps via already-merged commits (7163081 and others) that didn't rebaseline after growing them. Documented and bumped in file-size-baseline.json. - chat-helpers.test.ts: two gpt-5.5 routing assertions predate #9275 (fix(routing): bare model ids route to codex first), which deliberately made gpt-5.5 route to codex unconditionally, regardless of which other providers are active. Confirmed via #9275's own commit message and code comments this is intentional, not a regression; verified reproducible on the raw base tip alone, with no changes from this PR involved. Updated both assertions and their names to match the new, intentional default. * ci: re-trigger checks after GitHub Actions incident (2026-08-07, resolved) * ci: re-trigger checks (previous push event was dropped) * fix(quality): rebaseline combo-routing-engine.test.ts own-comment growth The ALL_ACCOUNTS_INACTIVE->ALL_TARGETS_SKIPPED fix (a32aed7) added explanatory comments (+7 lines), pushing the file past its frozen 3457 cap. CI's PR-mode check:file-size caught it; local check-file-size.mjs was not re-run after that specific commit.

CODEX_NATIVE_UNPREFIXED_MODELS lists the Codex model ids that should reach the ChatGPT subscription when a client sends them without a provider prefix (diegosouzapw#9275), bounded to installs with an active Codex connection (diegosouzapw#9447). GPT-6 Astra came to the Codex catalog in diegosouzapw#13026 with the same effort tiers as gpt-5.6-sol/terra, but its ids were never added to the set. With both a Codex and an OpenAI connection active: - bare `gpt-5.6-sol` routed to Codex, bare `gpt-6-astra` routed to OpenAI; - /v1/models listed bare `gpt-5.6-*` ids but no bare `gpt-6-astra*` id. Add gpt-6-astra and its six effort variants. The diegosouzapw#9447 bound still applies: OpenAI's catalog also lists `gpt-6-astra`, so without an active Codex connection the bare id keeps resolving to OpenAI. The effort variants are Codex-family ids (codex and codex-app-server); they already routed to Codex when it was active, so for them only the /v1/models rows change, exactly as for the gpt-5.6 variants. Test: tests/unit/codex-gpt6-astra-bare-id.test.ts (3 of 4 cases fail on the previous model.ts; the fourth pins the no-Codex fallback). Three comments in models-catalog-low-noise-flag.test.ts counted the set's size ("26 of the 27") and one cited its old line range; they now describe the one exception without a count and point at the file only.

…ndidates (diegosouzapw#9275) * fix(routing): bare model ids route to codex first; validate synced candidates Two bare-model-routing bugs surfaced in the field when an OmniRoute deployment had a codex subscription whose cookie quota was exhausted (retry-after 429047s / ~5 days) AND an active kiro connection whose upstream sync briefly advertised 'claude-opus-5' before kiro vendored it into the static registry. 1. Bare 'gpt-5.6-sol' (and friends) routed to the codex provider even when the user had explicitly configured 'agentrouter' as their provider (via model_provider in codex CLI). With codex in cooldown, every bare request 429'd. Fix: extend CODEX_NATIVE_UNPREFIXED_MODELS to include the full gpt-5.6-sol tier set + gpt-5.5 + the related codex-native ids. The Codex CLI default is now actually honored; users can still prefix 'agentrouter/gpt-5.6-sol' to opt into a specific provider. 2. Bare 'claude-opus-5' silently routed to 'kiro' when kiro's synced /v1/models catalog had that id (likely from a transient upstream quirk). kiro's static registry never cataloged claude-opus-5, so the upstream call 404'd. Fix: validate activeSyncedProviders against MODEL_TO_PROVIDERS before merging them into the candidate list. Auto-discovery still wins when the model id has no static entry (brand-new models from upstream keep working). Bonus: when handleNoCredentials returns a 404 'No active credentials for provider: X' error, surface the top-3 candidate aliases (e.g. 'anthropic/claude-opus-5, claude/claude-opus-5, agentrouter/claude-opus-5') so the operator can pick a working prefix instead of staring at a wall. Tests (all pass, 25 regression tests preserved): - tests/unit/fix-bare-model-precedence.test.ts (7 tests) - tests/unit/fix-synced-model-validation.test.ts (3 tests) - tests/unit/fix-error-message-candidates.test.ts (3 tests) - tests/unit/fix-bare-routing-fallback.test.ts (7 tests) * fix(tests): replace lorem ipsum with neutral text to avoid agentrouter WAF The agentrouter.org WAF blocks requests containing 'lorem ipsum' in messages[].content. When Claude Code reads test files via the Read tool, the content appears in tool_result blocks which can trigger the filter. Replace 'lorem ipsum dolor sit amet' with 'example content for testing purposes' in compression harness test to avoid false positives. --------- Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

…uzapw#9323) The agentrouter.org upstream WAF returns 400 content-blocked intermittently when: 1. messages[].content contains a blocked keyword (Lorem ipsum, the phrase 'language model' alone, 'virtual assistant', etc.); or 2. Requests from the same IP/key arrive in a burst, after which the WAF's per-IP suspicion bucket starts blocking content that would normally pass. The bucket relaxes after ~5-10s of idle. Apply three mitigations: 1. Burst guard (open-sse/services/wafRateLimit.ts) Per-bucket (provider+url) gate that enforces a 500ms minimum gap between outbound requests to agentrouter. Configurable via configureWafRateLimit(). Tested in tests/unit/wafRateLimit.test.ts. 2. Reactive retry (BaseExecutor.WAF_RETRY_CONFIG in base.ts) New WAF_RETRY_CONFIG with maxAttempts=2, delayMs=1500, backoffMultiplier=2. When the upstream returns 400 with a body that matches /content[_-]blocked/i, retry the same URL with exponential backoff (1.5s, 3.0s) before falling through to the 429/401/fallback chain. Tested in tests/unit/base-executor-waf-retry.test.ts. 3. Documentation (docs/security/AGENTROUTER_WAF.md) Blocklist of always-blocked and almost-always-blocked patterns, behavior under load, guidance for prompts/tool output, and pointers to the relevant code paths in OmniRoute. These are belt-and-suspenders: the burst guard prevents the WAF from activating on normal traffic, and the reactive retry recovers when it does anyway. Together they should eliminate the intermittent 400 content-blocked that Claude Code sees when running through agentrouter via OmniRoute. Refs diegosouzapw#9275 follow-up. Test: 'WAF retry config shape' and 'WAF retry differs from generic' guard the WAF_RETRY_CONFIG contract so future refactors don't accidentally collapse the two retry paths. Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

…codex is active (diegosouzapw#9447) * fix(routing): only let Codex-native bare ids preempt a provider when codex is active diegosouzapw#9275 widened CODEX_NATIVE_UNPREFIXED_MODELS from a single id to gpt-5.5 plus the gpt-5.6-sol/terra/luna tiers, so bare Codex CLI ids would reach the ChatGPT subscription instead of fanning out to whichever provider won the inference race. The early return it added never consulted the active-provider set, which made the codex-only guard 30 lines below unreachable for every id in the set: if (CODEX_NATIVE_UNPREFIXED_MODELS.has(modelId)) return { provider: "codex", ... } An OpenAI-only install therefore had bare gpt-5.5 routed to codex and failed with 'no active credentials for provider: codex' on a model OpenAI serves, and an install whose codex connection was merely inactive failed identically. This also silently reverted diegosouzapw#5887's compatibility boundary. The preference now only PREEMPTS another provider when a codex connection is active. Ids that no other provider catalogs (codex-auto-review) still resolve to codex with no connection at all — there is nothing to preempt and 'no codex credentials' is the honest error. With codex active the preference still beats OpenAI, which is the point of diegosouzapw#9275, and an explicit openai/ prefix overrides it either way. Tests: the three assertions that encode the intended diegosouzapw#9275 change now expect codex (plus a new one pinning the explicit-prefix override); the rest were already correct and pass again untouched. Adds a regression test for the OpenAI-only case. * docs(changelog): correct fragment id to diegosouzapw#9447 * test(routing): seed an active codex connection in the bare-precedence guards The two files diegosouzapw#9275 added assert that bare gpt-5.5 / gpt-5.6-sol reach codex, but they ran against an empty database — so they also pinned 'codex wins with no codex connection at all', which is the regression diegosouzapw#9447 removes. That put them in direct contradiction with plan3-p0 / chat-helpers / codex-gpt55-routing-5887, which assert openai for the very same input: no implementation could satisfy both, which is why the release could not go green. Seeding an active codex connection keeps the contract these files were written to guard (codex beats openai for a Codex-native bare id) while dropping the accidental 'even with no codex configured' half. Cases that need no connection are left as they were: the tier-only ids and codex-auto-review have no alternative provider to preempt, and the explicit-prefix overrides are unaffected. --------- Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

…iegosouzapw#9392) diegosouzapw#9275 started appending a candidate-alias hint to the zero-active-credentials error and terminated the provider name with a period, so the two sentences read as one message. The two vscode tokenized-route tests still assert the old unterminated string and now fail on every pull request opened against this branch. The Quality Gates workflow only runs on pull_request to release/**, never on push, so the branch itself never re-runs these shards and the drift stayed invisible after the merge. Assert what the handler actually produces. Keeping the comparison exact rather than loosening it to a prefix match is deliberate -- the exact form is what caught the drift. Signed-off-by: Minxi Hou <houminxi@gmail.com>

…n every provider (diegosouzapw#9006) * fix(executors): route Claude-via-Vertex through native rawPredict with real streaming Claude models on Vertex AI were being sent through the generic OpenAI- compatible partner endpoint, which 404s/errors for Claude on at least some projects. Route them through Vertex's native Anthropic Messages API (publishers/anthropic/.../rawPredict) instead, stripping the body-level model field rawPredict rejects and injecting the required anthropic_version field. rawPredict only ever returns a complete JSON body, never real SSE framing, so streaming requests now get a genuine Anthropic-format SSE stream synthesized from that JSON (message_start/content_block_*/ message_delta/message_stop), which the existing claude-to-openai response translator already knows how to parse. Also fixes two response-format resolution bugs that silently dropped a custom model's DB-stored targetFormat override whenever the model id also existed in the static provider registry (as claude-sonnet-4-6 and claude-opus-4-7 do under vertex): resolveModelOrError had its own ad-hoc resolution that never consulted the override, and even once fixed, executeChatWithBreaker discarded the correctly-resolved format before handleChatCore's own resolution ran a second time. * docs: add changelog fragment for diegosouzapw#8909 * refactor(sse): extract shared Claude effort-model predicate * fix(sse): strip Claude effort-suffix ids for any provider serving a real Claude model * fix(sse): keep no-think and CC-discovery catalog variant roots unprefixed * fix(dashboard): re-qualify no-think playground model ids correctly * fix(sse): scope Vertex 404s to a per-model lockout via passthroughModels * docs: add changelog fragment for the Claude catalog/dispatch fix * fix(sse): align regex naming and changelog formatting * fix(sse): clarify effort-variant strip comment and add cross-module drift guard * fix(sse): disambiguate Vertex connection-wide vs per-model 403s * docs: document Vertex 403 disambiguation in changelog fragment * fix(sse): correlate reason and resource within the same ErrorInfo detail * fix(sse): extract Vertex error classifier and rebaseline frozen file sizes * test: register vertex-passthrough-model-lockout in stryker tap.testFiles * fix(sse): reconciles rebase-onto-tip drift for 9006 Two categories of inherited base-branch breakage surfaced when rebasing onto release/v3.8.50's latest tip, both confirmed unrelated to this PR's own diff: - check:file-size: base.ts and chat.ts drifted further past their frozen caps via already-merged commits (174e08b and others) that didn't rebaseline after growing them. Documented and bumped in file-size-baseline.json. - chat-helpers.test.ts: two gpt-5.5 routing assertions predate diegosouzapw#9275 (fix(routing): bare model ids route to codex first), which deliberately made gpt-5.5 route to codex unconditionally, regardless of which other providers are active. Confirmed via diegosouzapw#9275's own commit message and code comments this is intentional, not a regression; verified reproducible on the raw base tip alone, with no changes from this PR involved. Updated both assertions and their names to match the new, intentional default. * ci: re-trigger checks after GitHub Actions incident (2026-08-07, resolved) * ci: re-trigger checks (previous push event was dropped) * fix(quality): rebaseline combo-routing-engine.test.ts own-comment growth The ALL_ACCOUNTS_INACTIVE->ALL_TARGETS_SKIPPED fix (a32aed7) added explanatory comments (+7 lines), pushing the file past its frozen 3457 cap. CI's PR-mode check:file-size caught it; local check-file-size.mjs was not re-run after that specific commit.

diegosouzapw added 2 commits August 3, 2026 02:48

diegosouzapw merged commit a72e165 into release/v3.8.50 Aug 3, 2026
10 checks passed

diegosouzapw mentioned this pull request Aug 3, 2026

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

Merged

HouMinXi mentioned this pull request Aug 4, 2026

test(sse): expect the trailing period in the no-credentials message #9392

Merged

wgordon17 mentioned this pull request Aug 4, 2026

fix(sse): restores #5887 openai precedence for bare gpt-5.5 routing #9440

Closed

diegosouzapw deleted the fix/bare-model-routing-precedence branch August 4, 2026 20:50

QuangBlue mentioned this pull request Sep 24, 2026

fix(routing): treat bare GPT-6 Astra ids as Codex-native #14721

Merged

Navigation Menu

Uh oh!

fix(routing): bare model ids route to codex first; validate synced candidates - #9275

fix(routing): bare model ids route to codex first; validate synced candidates#9275
diegosouzapw merged 2 commits into
release/v3.8.50from
fix/bare-model-routing-precedence

diegosouzapw commented Aug 3, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

diegosouzapw commented Aug 3, 2026

Files changed

Validation

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant