Skip to content

chore(release): clear release/v3.8.51 base-red gates — docs count, stryker list, lockfile host, stale suppressions, 7 lint regressions - #11502

Merged
diegosouzapw merged 7 commits into
diegosouzapw:release/v3.8.51from
MumuTW:fix/release-v3.8.51-basereds
Aug 25, 2026
Merged

diegosouzapw merged 7 commits into
diegosouzapw:release/v3.8.51from
MumuTW:fix/release-v3.8.51-basereds

Conversation

@MumuTW

@MumuTW MumuTW commented Aug 25, 2026 •

Copy link
Copy Markdown
Contributor

Part of #11449. Every fix here reproduces on the pristine release/v3.8.51 tip with no other changes, and none alters runtime behaviour.

What turns green

check cause on base fix
Docs Gates (check:fabricated-docs strict drift) README.md / AGENTS.md / llm.txt say "159 migrations", code has 160 count updated in the three root docs and the 38 strict i18n llm.txt mirrors (check:docs-sync requires byte-identical bodies, so root-only edits red it)
Fast Quality Gates → check:mutation-test-coverage --strict tests/unit/repro-combo-persisted-cooldown-preskip.test.ts and tests/unit/repro-glm-iso-reset-24h-cap.test.ts cover mutated modules but are absent from stryker.conf.json tap.testFiles both added, in place
Fast Quality Gates → check:lockfile libxmljs2/node_modules/brace-expansion resolved from registry.npmmirror.com host → registry.npmjs.org; same tarball, same integrity hash
No new ESLint warnings (exit 2, before any lint result) two stale entries in config/quality/eslint-suppressions.json (videoBridgeContactSheet.ts, videoBridgeRuntime.ts) eslint --prune-suppressions
No new ESLint warnings (275 errors, visible only once exit 2 is gone) #11434 orphaned windowStillExhaustedAfterRealReset() + 3 imports in providerLimits.ts and 1 import in videoBridge.ts; #11390's new test added an unused payload and two as any, taking combo-routing-engine.test.ts to 269 vs a 267 bulk-suppression quota (all-or-nothing → all 269 surface) orphaned code deleted (zero references); relayOptions already satisfies ComboRelayOptions, cast dropped; file back at exactly 267
Unit fast-path → providers-constants-split.test.ts #11434 added volcengine-agent-plan + volcengine-coding-plan (regional) without bumping the hard-coded partition count assertion 231 → 233 (233 keys, 233 distinct, family sum 233); history comment extended in the file's own style
Build (advisory) + hook typecheck + live-ws-url-11331.test.ts #11388 (merged 08-24) added sanitizeLiveWsPort / resolveLiveWsUrl to wsPath.ts and wired them into useLiveDashboard.ts; the next day's #11434 carried a competing #11331 fix whose wsPath.ts overwrote that hunk while hook + test kept the imports → TS2305, red test, Turbopack "Export resolveLiveWsUrl doesn't exist" the two functions re-added verbatim beside the surviving resolveLiveWsPublicUrl; 11/11, tsc clean
Vitest (fast-path) → glmCodingProviderConfig.test.ts registry ships glm-5.3-max with a max tier; Z.AI's GLM-5.3 reference documents reasoning_effort ∈ {low, high, max} (max recommended for coding), so the test's inventory/tier tables were stale glm-5.3-max + ["max"] added; 10/10
check:open-sse-typecheck catalog.ts:1614,1648 — `!modelType

Verification (local, this branch)

npm run check:docs-sync                 → PASS, 42 locales match
npm run check:mutation-test-coverage    → ✓ No drift
npm run check:lockfile                  → OK, no issues
npm run lint:json -- --max-warnings 0   → exit 0
npm run typecheck:core                  → clean
node --test tests/unit/combo-routing-engine.test.ts   → 91/91
node --test <12 unit files importing providerLimits / videoBridge> → 73/73
node --test tests/unit/providers-constants-split.test.ts → 4/4

Deliberately NOT touched (need the owning author's judgement)

  • Unit fast-path shards — the 14 test files behind the remaining ✖ lines were run on the pristine base tip 6435f618f4 and on this PR's tip b4bb5ba8bb with identical results: 100 tests, 19 failures, the same 19 on both (A2A agent-card skill count ×5, conductor agent-card ×2, provider-nodes CC mode ×2, postExchange onboarding ×4, openapi coverage floor, route-body Zod t06, pack-artifact closures, hard-lease inventory, translate-path golden, minimax-m3 ledger, Claude/GLM quota recovery ×2). Three names that show only on CI (buildWeeklyQuotaFallback…, checkFallbackError… GLM 429, Antigravity BYOP…) pass locally on both trees — timing flakes, not regressions.

Those are listed on #11449 with reproduction notes.

MumuTW added 3 commits August 25, 2026 18:40
…stryker tap.testFiles, lockfile host, stale suppressions

Four mechanical fixes for gates that fail on the pristine release/v3.8.51 tip
(tracked in diegosouzapw#11449); none change runtime behaviour.

- docs: "159 migrations" → "160" in README.md, AGENTS.md, llm.txt and the 38
  strict i18n llm.txt mirrors (check:docs-sync requires byte-identical bodies).
- stryker.conf.json: add tests/unit/repro-combo-persisted-cooldown-preskip.test.ts
  and tests/unit/repro-glm-iso-reset-24h-cap.test.ts to tap.testFiles; they cover
  mutated modules and check:mutation-test-coverage --strict flags the drift.
- package-lock.json: libxmljs2/node_modules/brace-expansion resolved from
  registry.npmmirror.com → registry.npmjs.org (same tarball, same integrity);
  check:lockfile only allows the canonical host.
- config/quality/eslint-suppressions.json: prune two stale entries
  (videoBridgeContactSheet.ts, videoBridgeRuntime.ts) that made ESLint exit 2
  before reporting anything.
…uning stale suppressions

Once the two stale suppression entries are pruned (previous commit), ESLint stops
exiting 2 and reports 275 real errors on the base tip — all introduced today:

- src/lib/usage/providerLimits.ts: diegosouzapw#11434 removed the callers of
  windowStillExhaustedAfterRealReset() (added in diegosouzapw#10534) and of three
  claudeExtraUsage imports but left the definitions behind. Delete the orphaned
  function and imports; no behaviour change (zero references).
- src/lib/guardrails/videoBridge.ts: unused formatVideoTimestamp import (diegosouzapw#11434).
- tests/unit/combo-routing-engine.test.ts: the quota-share test added in diegosouzapw#11390
  introduced an unused `payload` and two `as any` casts, pushing the file to 269
  no-explicit-any against a bulk-suppression quota of 267; ESLint's suppression
  is all-or-nothing, so every one of the 269 surfaced. `relayOptions` already
  satisfies ComboRelayOptions (all fields optional), so the cast is dropped and
  the unused line removed; the file is back at exactly 267.

Verification: `npm run lint:json -- --max-warnings 0` exit 0;
`npm run typecheck:core` clean; combo-routing-engine.test.ts 91/91; the 12 unit
files importing providerLimits/videoBridge 73/73.
MumuTW added 4 commits August 25, 2026 19:01
…11434 added the two volcengine plans

diegosouzapw#11434 added volcengine-agent-plan and volcengine-coding-plan to the regional family
without bumping the partition assertion in providers-constants-split.test.ts, which
hard-codes the merged count. Actual: 233 keys, 233 distinct, family sum 233.
…rom wsPath.ts; add glm-5.3-max to the GLM test table

Two more release/v3.8.51 base-reds (diegosouzapw#11449), both mechanical once traced:

- src/shared/utils/wsPath.ts: diegosouzapw#11388 (merged 2026-08-24) fixed diegosouzapw#11331 by adding
  sanitizeLiveWsPort() + resolveLiveWsUrl() and wiring them into
  useLiveDashboard.ts with tests/unit/live-ws-url-11331.test.ts. The next day's
  diegosouzapw#11434 merge carried a competing diegosouzapw#11331 fix (resolveLiveWsPublicUrl) and its
  wsPath.ts version overwrote diegosouzapw#11388's hunk, while the hook and the test kept
  importing the two functions. Result on the tip: TS2305 in the hook, the unit
  test red, and the Turbopack build failing on "Export resolveLiveWsUrl doesn't
  exist". This re-adds diegosouzapw#11388's two functions verbatim next to the surviving
  resolveLiveWsPublicUrl; the names do not collide and the precedence they
  implement (explicit wsUrl → handshake publicUrl → default URL with handshake
  port/path) is unchanged.

- open-sse/mcp-server/__tests__/glmCodingProviderConfig.test.ts: the registry
  ships glm-5.3-max with a "max" effort tier. Z.AI's GLM-5.3 reference lists
  reasoning_effort ∈ {low, high, max} and recommends max for coding, so the
  registry is right and the inventory/tier tables in the test were stale (they
  already declared ["low","high","max"] on the base model).

Verification: tests/unit/live-ws-url-11331.test.ts 11/11; glmCodingProviderConfig
10/10; tsc reports no errors in useLiveDashboard.ts / wsPath.ts.
… as TS2367 on release/v3.8.51

Before diegosouzapw#11436, catalog.ts derived `modelType` as a local `string | undefined` that
was only ever assigned "embedding" | "rerank" | "image" | "audio" — "chat" was
never produced, and chat models were represented by `undefined`. diegosouzapw#11436 replaced
that block with classifyModelSupportedEndpoints(), whose literal return union
makes the same comparison a TS2367 (baseline 0 → live 2) and reds
check:open-sse-typecheck. `!modelType || modelType === "chat"` is therefore
exactly `!modelType`; behaviour is unchanged at both sites.

Verification: npm run check:open-sse-typecheck → OK (5 pre-existing errors, all
within the frozen baseline); the 8 unit files importing the catalog route 61/61.
@diegosouzapw
diegosouzapw merged commit 17e4ddf into diegosouzapw:release/v3.8.51 Aug 25, 2026
12 of 16 checks passed
diegosouzapw pushed a commit that referenced this pull request Aug 25, 2026
Validated in a combined 3-PR batch worktree off release/v3.8.51 tip. The prerequisite sibling (159→160 migrations) was already covered by #11502 — closed the redundant #11543 separately.
- New check:docs-counts gate: proven fail-before/pass-after per the PR's own methodology, re-verified here (PASS, 2 unrelated soft-drift notices on cloud-agent/A2A doc counts, non-blocking)
- Focused tests: check-docs-counts-sync.test.ts, combo-matrix/auto.test.ts, lkgp-enabled-context-11181.test.ts — pass
- typecheck:core, file-size, changelog-integrity, complexity, cognitive-complexity — all OK

Thanks for replacing seventeen hand-maintained factor-count claims with one source-derived gate.
diegosouzapw pushed a commit that referenced this pull request Aug 26, 2026
#11484)

Validated in a combined sub-batch worktree off release/v3.8.51 tip. Its second commit reimplemented wsPath.ts's sanitizeLiveWsPort/resolveLiveWsUrl to fix a build break it hit — that exact break is already fixed on the current tip via #11502 (the original #11388 implementation restored verbatim). Conflicted against that; resolved by keeping the tip's established implementation and dropping the redundant reimplementation (both are functionally equivalent; the tip's carries the original #11331 doc comments and precedence contract). Pushed the same resolution to this branch.
- typecheck:core, file-size, changelog-integrity, complexity, cognitive-complexity — all OK
- Full-repo lint: 228 pre-existing dashboard react-hooks/* findings, unrelated to this diff
- Focused test: cli-settings-jcode.test.ts (integration) — part of sub-batch's 165/165 run

Thanks for the real end-to-end verification (production server, real ~/.jcode/config.toml, jcode --provider-profile omniroute serving a completion, jcode provider-doctor passing) — the managed-block approach (marker-delimited, preserves user-authored config, 409s instead of corrupting on a hand-written conflicting table) is exactly the right shape for this integration.
@MumuTW
MumuTW deleted the fix/release-v3.8.51-basereds branch August 27, 2026 01:47
muhamadgalihsaputra pushed a commit to niyatna/NiyatnaRoute that referenced this pull request Sep 27, 2026
…ryker list, lockfile host, stale suppressions, 7 lint regressions (diegosouzapw#11502)

Validated in a combined 4-PR batch worktree off release/v3.8.51 tip.
- Every fix individually confirmed against the pristine tip, no runtime behavior change
- typecheck:core, file-size, changelog-integrity, complexity, cognitive-complexity, check:cycles — all OK
- Full-repo lint: 503 → 228 problems, confirming this PR's diagnosis of the exit-2 stale-suppressions + orphaned-code causes; the remaining 228 are pre-existing dashboard react-hooks/* findings this PR never claimed to touch
- node --test tests/unit/combo-routing-engine.test.ts, providers-constants-split.test.ts, and the providerLimits/videoBridge importers — all pass as part of the batch's 246/246 node:test run

Thanks for the meticulous base-red triage — this directly explains and fixes the largest lint-drift finding from the prior merge-batch session.
muhamadgalihsaputra pushed a commit to niyatna/NiyatnaRoute that referenced this pull request Sep 27, 2026
…osouzapw#11545)

Validated in a combined 3-PR batch worktree off release/v3.8.51 tip. The prerequisite sibling (159→160 migrations) was already covered by diegosouzapw#11502 — closed the redundant diegosouzapw#11543 separately.
- New check:docs-counts gate: proven fail-before/pass-after per the PR's own methodology, re-verified here (PASS, 2 unrelated soft-drift notices on cloud-agent/A2A doc counts, non-blocking)
- Focused tests: check-docs-counts-sync.test.ts, combo-matrix/auto.test.ts, lkgp-enabled-context-11181.test.ts — pass
- typecheck:core, file-size, changelog-integrity, complexity, cognitive-complexity — all OK

Thanks for replacing seventeen hand-maintained factor-count claims with one source-derived gate.
muhamadgalihsaputra pushed a commit to niyatna/NiyatnaRoute that referenced this pull request Sep 27, 2026
diegosouzapw#11484)

Validated in a combined sub-batch worktree off release/v3.8.51 tip. Its second commit reimplemented wsPath.ts's sanitizeLiveWsPort/resolveLiveWsUrl to fix a build break it hit — that exact break is already fixed on the current tip via diegosouzapw#11502 (the original diegosouzapw#11388 implementation restored verbatim). Conflicted against that; resolved by keeping the tip's established implementation and dropping the redundant reimplementation (both are functionally equivalent; the tip's carries the original diegosouzapw#11331 doc comments and precedence contract). Pushed the same resolution to this branch.
- typecheck:core, file-size, changelog-integrity, complexity, cognitive-complexity — all OK
- Full-repo lint: 228 pre-existing dashboard react-hooks/* findings, unrelated to this diff
- Focused test: cli-settings-jcode.test.ts (integration) — part of sub-batch's 165/165 run

Thanks for the real end-to-end verification (production server, real ~/.jcode/config.toml, jcode --provider-profile omniroute serving a completion, jcode provider-doctor passing) — the managed-block approach (marker-delimited, preserves user-authored config, 409s instead of corrupting on a hand-written conflicting table) is exactly the right shape for this integration.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants