fix(combo): do not treat credits-exhausted 401 as auth skip

[RaviTharuma]

Summary

• Combo dispatch treated HTTP 401/403 as authentication-expired (fix(resilience): round-robin combo doesn't skip circuit-broken provider (opencode-zen 401 cascade) #8133) even when the body was quota/credits exhaustion, so remaining combo targets were skipped.
• credits_exhausted is now surfaced as HTTP 402 (PAYMENT_REQUIRED) from chat and embeddings credential exhaustion, matching the quota class.
• applyComboTargetExhaustion ignores quota/credits bodies on the auth-skip path. Real expired 401 still marks the connection.

Related Issues

• Closes [BUG] Successful OAuth refresh and non-auth failures (quota, 403, combo timeout) are classified as authentication expired and skip remaining combo targets #12441
• Related to fix(resilience): round-robin combo doesn't skip circuit-broken provider (opencode-zen 401 cascade) #8133, [Feature] (combo): provider-level exhaustion tracking to skip same-provider targets #1731, fix(resilience): separate combo quality and authentication failure summaries #10314, fix(resilience): ALL_TARGETS_SKIPPED 503 on round-robin combos returns no per-target diagnostics #12294

Validation

• Change type: routing
• Focused tests and category gates from the golden path
• npm run lint (CI)
• Reconciled with the current active release base; focused checks rerun afterward
• Production-code changes include a new or updated automated test in this PR

Focused command:

node --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test tests/unit/12441-quota-not-auth-skip.test.ts tests/unit/chat-helpers.test.ts

Result: 32 pass, 0 fail.

Tests Added Or Updated

• tests/unit/12441-quota-not-auth-skip.test.ts
• tests/unit/chat-helpers.test.ts

Coverage Notes

• Combo skip classification is covered by the new [BUG] Successful OAuth refresh and non-auth failures (quota, 403, combo timeout) are classified as authentication expired and skip remaining combo targets #12441 unit tests.
• Chat credential exhaustion HTTP status is covered by handleNoCredentials maps credits_exhausted to HTTP 402 not 401.

Reviewer Notes

• This does not change genuine 403 auth skip (invalid token / embeddings 403).
• Combo per-model timeout vs upstream fetch timeout is still a follow-up from [BUG] Successful OAuth refresh and non-auth failures (quota, 403, combo timeout) are classified as authentication expired and skip remaining combo targets #12441.
• TOKEN_REFRESH success still does not by itself clear skip state; that is a separate follow-up.

[Copilot]

🟡 Changes recommended

The new quota/credits detection can miss quota signals when structured errors include a non-quota code, and the embeddings 402 mapping lacks a direct regression test.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

Fixes combo exhaustion classification so quota/credits exhaustion that surfaces as HTTP 401/403 does not take the auth-skip path (#8133), and surfaces credits_exhausted as HTTP 402 (Payment Required) from chat + embeddings to align with quota semantics.

Changes:

• Map credits_exhausted terminal credential state to HTTP 402 (chat + embeddings) instead of 401.
• Add isQuotaOrCreditsError() and use it to prevent misclassified quota/credits bodies from triggering auth-level combo exhaustion.
• Add unit tests for the #12441 combo skip regression and the chat helper HTTP status mapping.

Commands / Coverage (custom):

• Commands run (per PR description): node --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test tests/unit/12441-quota-not-auth-skip.test.ts tests/unit/chat-helpers.test.ts
• Changed test files: tests/unit/12441-quota-not-auth-skip.test.ts, tests/unit/chat-helpers.test.ts
• Coverage result (custom): Not provided/verified in this review (no npm run test:coverage output included).

File summaries

File	Description
open-sse/services/combo/targetExhaustion.ts	Adds quota/credits detection helper and blocks auth-skip for misclassified quota/credits error bodies.
src/sse/handlers/chatHelpers.ts	Maps credits_exhausted terminal state to HTTP 402 instead of 401 in handleNoCredentials.
src/lib/embeddings/service.ts	Maps embeddings credential exhaustion with credits_exhausted to HTTP 402.
tests/unit/chat-helpers.test.ts	Adds regression test asserting handleNoCredentials returns 402 for credits_exhausted.
tests/unit/12441-quota-not-auth-skip.test.ts	New unit tests covering quota/credits bodies not triggering auth-level combo skips.

Review details

• Files reviewed: 5/5 changed files
• Comments generated: 3
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[diegosouzapw]

sweep-reds / babysit: reds on this PR are inherited from base-red #12581, not this quota-skip diff. Skipping a code change; did not merge origin/release/v3.8.51.

• ESLint: CodeQL ratchet 12 open > baseline 11 (repo-wide Security alerts). No unused/@eloqnt in this diff.
• FQG mutation-test-coverage: missing tests/unit/video-bridge-memory-suppression.test.ts on open-sse/handlers/chatCore/memoryExtraction.ts — already listed on the current origin/release/v3.8.51 tip; not a file this PR touches.
• Unit Tests 3/4: tests/unit/build/npm-ci-retry-composite.test.ts (fixTlsClientNodeBinary.mjs vanished / wreqJsNative.mjs omitted). Not in this PR's files; already corrected on the release tip.

[diegosouzapw]

sweep-reds round 6 / babysit: re-verified — every remaining red is inherited from base-red #12581, not this quota-skip diff. Skipping a code change; did not merge origin/release/v3.8.51.

• ESLint: eslintWarnings=0; job fails CodeQL ratchet 12 open > baseline 11.
• FQG: mutation-test-coverage missing tests/unit/video-bridge-memory-suppression.test.ts covering memoryExtraction.ts — not this PR (targetExhaustion.ts / embeddings / chatHelpers.ts + 12441 tests). Complexity new-code was OK (3 files).
• Units 3/4: fixTlsClientNodeBinary.mjs vanished / wreqJsNative.mjs cache-key — postinstall helpers this PR does not touch.

STOP. Resume after #12581 drains.