fix(review): rate-limit the ci_stuck_review_repeat_suppressed log (#4998)

[JSONbored]

Summary

• Fixes fix(review): ci_stuck_review_repeat_suppressed logs itself as an error on every suppression (650 Sentry events) #4998: ci_stuck_review_repeat_suppressed — a log line that exists specifically to announce "I am suppressing a repeat notification" — fired as its own console.error (Sentry-forwarded) on every single suppression: 649 events over 4 days, escalating.
• Root cause matched Sentry Seer's summary directly this time: the #orb-ci-stuck-repeat guard (prReadyForReview, src/queue/processors.ts) correctly caps the expensive part (re-running a full paid AI review) to once per head SHA, but the log announcing that cap re-fires on every later evaluation of the same still-stuck PR — a webhook re-trigger or a sweep pass, at roughly a 5-10 minute cadence based on the raw event timestamps.
• Fix: rate-limit the log only, once per (repo, pr, headSha) per day, via the same SELFHOST_TRANSIENT_CACHE-backed helper pattern already used by ciPendingDeferStuck/ciReReviewCoalesced in this file (getTransientKey/putTransientKey). The underlying defer/suppression behavior is untouched — it still runs on every evaluation; only the Sentry-visible log is coalesced. A new head SHA (a fresh commit) is a fresh key, so a genuinely new incident is never silently swallowed by an old PR's coalesce window.

Investigation: what's actually wrong with the stuck PR(s) (issue requirement #2)

Pulled the raw event list behind this Sentry issue rather than trusting its title/culprit summary, since forwardStructuredLogToSentry groups every ci_stuck_review_repeat_suppressed occurrence into one Sentry issue regardless of repo/PR/headSha (grouped by the fixed message template) — the issue's displayed "(JSONbored/awesome-claude#4816) headSha=..." is just the latest contributing event, not a single PR stuck for 4 days as the issue title implied. Confirmed: PR awesome-claude#4816 itself was only open ~77 minutes end-to-end (created 04:44 UTC, merged 06:01 UTC the same day) — not stuck at all.

The real 649-event history spans several distinct PRs across gittensory and awesome-claude (at least #4343, #4344, #4346, #4349, #4371 in gittensory and #4739, #4745, #4816 in awesome-claude, from a 25-event sample), each re-evaluated repeatedly at a ~5-10 minute cadence while stuck — the exact "steady few-minute cadence" burn pattern the original #orb-ci-stuck-repeat guard's own code comment describes, just now visible as noisy logging instead of wasted AI spend.

Checked the most-repeated one, gittensory#4349, directly: it carries a non-required Contributor trust check (a third-party Action, superagent.sh) that resolved to conclusion ACTION_REQUIRED in 3 seconds — but main's actual branch-protection required-status-checks list is only ["Superagent Security Scan", "validate"], so that check is not itself the blocker. All sampled PRs eventually merged cleanly once a later commit landed or a required check finally completed, so this is not a permanent hang — it's multiple PRs sitting through a slow-to-settle CI run (or a required check needing a manual fork-Action approval) for anywhere from tens of minutes to a few hours each, during which the sweep/webhook cadence re-evaluates and re-logs. No code change beyond this PR's rate-limit is proposed for that underlying CI-settling latency; it's a candidate follow-up (documented in the parent observability epic) rather than a fix bundled here.

Scope

• The PR title follows type(scope): short summary Conventional Commit format, for example fix(api): restore profile access checks.
• This PR is focused and does not mix unrelated backend, UI, MCP, docs, dependency, and deploy changes.
• This follows CONTRIBUTING.md and does not reintroduce GitHub Pages, VitePress, site/, or CNAME.
• I linked a currently open issue this PR resolves (Closes #4998).

Validation

• git diff --check
• npm run actionlint
• npm run typecheck
• npm run test:coverage locally
• npm run test:workers
• npm run build:mcp
• npm run test:mcp-pack
• npm run ui:openapi:check
• npm run ui:lint
• npm run ui:typecheck
• npm run ui:build
• npm audit --audit-level=moderate
• New or changed behavior has unit/integration tests for new branches, fallback paths, and sanitizer boundaries

If any required check was skipped, explain why:

• test:coverage (full unsharded): not run end-to-end — ran scoped vitest --coverage for test/unit/queue.test.ts (159 tests) and confirmed via lcov that every changed line and both branches of the new coalescing guard are covered.
• actionlint / test:workers / build:mcp / test:mcp-pack / ui:openapi:check / ui:lint / ui:typecheck / ui:build / npm audit: not run — this change touches only src/queue/processors.ts (an existing transient-cache-backed guard, no new API/schema/binding/dependency surface) and its tests; no workflow, MCP, UI, or dependency-manifest surface changed.

Safety

• No secrets, wallet details, hotkeys, coldkeys, user PATs, private keys, raw trust scores, private rankings, or private maintainer evidence are exposed.
• Public GitHub text stays sanitized, low-noise, and does not imply compensation guarantees or optimization tactics.
• Auth, cookie, CORS, GitHub App, Cloudflare, or session changes include negative-path tests. (N/A — no auth surface touched.)
• API/OpenAPI/MCP behavior is updated and tested where needed. (N/A.)
• UI changes use live API data or real empty/error/loading states. (N/A.)
• Visible UI changes include a UI Evidence section. (N/A.)
• Public docs/changelogs are updated where needed. (N/A — internal engine behavior; changelog is not edited in a normal PR.)

Notes

Part of a batch of 13 bug fixes filed from a Sentry-issue triage this session (#4994–#5006). This is #5 by priority.

[superagent-security]

Superagent didn't find any vulnerabilities or security issues in this PR.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 94.36%. Comparing base (5b2334e) to head (079d367).
⚠️ Report is 1 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5037   +/-   ##
=======================================
  Coverage   94.36%   94.36%           
=======================================
  Files         459      459           
  Lines       39285    39291    +6     
  Branches    14329    14331    +2     
=======================================
+ Hits        37072    37078    +6     
  Misses       1558     1558           
  Partials      655      655           

Flag	Coverage Δ
shard-1	46.45% <100.00%> (-0.28%)	⬇️
shard-2	33.63% <14.28%> (+0.25%)	⬆️
shard-3	31.36% <14.28%> (-0.12%)	⬇️
shard-4	32.57% <14.28%> (-0.26%)	⬇️
shard-5	34.34% <14.28%> (+0.88%)	⬆️
shard-6	45.41% <14.28%> (+0.30%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/queue/processors.ts	95.39% <100.00%> (+<0.01%)	⬆️

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.