fix(review): repair stale gate-check publication state

[JSONbored]

Summary

• Fix stale AI-review cache inputs so review results are keyed by the current review context.
• Persist successful Gate check publication into local check summaries.
• Repair scheduled sweep recovery so all current-head PRs missing the required Gate check are re-reviewed, even when ordinary stale maintenance is backpressured.

What changed

• Adds a dedicated AI review cache input fingerprint module and tests.
• Records local Gittensory Orb Review Agent summaries after publishing Gate checks.
• Treats missing current Gate checks as repair-priority sweep candidates.
• Lets repair-priority candidates bypass the webhook freshness skip and expand beyond the normal stale-maintenance cap.
• Keeps the normal low sweep cap and backlog deferral for ordinary stale maintenance.
• Adds regressions for cache invalidation, Gate summary persistence, missing-check repair priority, backlog bypass for repair, and healthy fresh PR no-op behavior.

Why

Required Gate checks could remain in GitHub as Expected — Waiting for status to be reported when local publication state was stale or a webhook was missed. The scheduler also repaired only a small stale batch per pass, so one repo could leave current-head required statuses stranded behind unrelated maintenance backlog.

Validation

• npx vitest run test/unit/agent-sweep.test.ts test/unit/queue.test.ts test/unit/ai-review-cache.test.ts
• npm run typecheck
• git diff --check
• Scoped coverage run completed all focused tests; changed-source LCOV check found no missed coverable changed lines and no changed-branch misses in the touched source files.

Notes

• The normal stale sweep cap remains in place for rate-limit safety. Only missing current Gate checks use the expanded repair batch.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.78%. Comparing base (385990b) to head (ede5fa1).
⚠️ Report is 3 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2351      +/-   ##
==========================================
+ Coverage   95.73%   95.78%   +0.05%     
==========================================
  Files         223      224       +1     
  Lines       24755    24917     +162     
  Branches     8970     9056      +86     
==========================================
+ Hits        23699    23867     +168     
+ Misses        433      428       -5     
+ Partials      623      622       -1     

Files with missing lines	Coverage Δ
src/db/repositories.ts	96.30% <100.00%> (+<0.01%)	⬆️
src/queue/processors.ts	91.07% <100.00%> (+0.85%)	⬆️
src/review/ai-review-cache-input.ts	100.00% <100.00%> (ø)
src/settings/agent-sweep.ts	100.00% <100.00%> (ø)

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

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

[loopover-orb]

Warning

🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨

⏸️ Gittensory review result - manual review recommended

_{Review updated: 2026-07-01 19:21:29 UTC}

8 files · 1 AI reviewer · no blockers · readiness 93/100 · CI green · clean

⏸️ Suggested Action - Manual Review

• Touches a guarded path — held for manual review

Review summary
The cache fingerprinting change correctly stops reusing AI-review rows when prompt/config inputs change, and the sweep now has a clear repair path for missing current Gate checks. The highest-risk regression is in public-surface completion: a successful Gate check is now counted as a published output, so configured comment/label failures can be stamped as complete and hidden from later repair. The backlog bypass also has an edge case where non-actionable priority PRs can enlarge the repair cap and pull ordinary stale work into a draining backlog.

Blockers

• src/queue/processors.ts:5813 and src/queue/processors.ts:5135 treat a published Gate check as sufficient output for finishPublicSurfacePublication, so in a comment_only or label_only repo where the Gate check succeeds but the configured comment/label write fails, markPullRequestSurfacePublished still stamps the head and the sweep will no longer repair the missing configured surface.
• src/queue/processors.ts:1350 computes repairCandidateLimit from raw priorityPullNumbers.length before src/settings/agent-sweep.ts:72 filters drafts, so an open draft that is marked priority can inflate the backlog-bypass cap and let an ordinary stale PR enqueue while per-PR backlog is active.

Nits — 6 non-blocking

• nit: src/review/ai-review-cache-input.ts:68 preserves changedPaths order in the fingerprint, so equivalent file sets can miss the cache if the upstream file listing order changes.
• nit: src/queue/processors.ts:5813 duplicates the same recordPublishedGateCheckSummary catch-and-log block across the primary and fallback Gate publication paths, which makes future changes easy to apply inconsistently.
• In src/queue/processors.ts:5135, make finishPublicSurfacePublication require all configured non-Gate public outputs to succeed before stamping lastPublishedSurfaceSha, or keep Gate publication out of publishedOutputs and track it separately.
• In src/queue/processors.ts:1350, derive the backlog repair cap from PRs that selectRegateCandidates can actually enqueue, such as open non-draft priority PRs, rather than from the raw priority number set.
• In src/review/ai-review-cache-input.ts:68, sort changedPaths before hashing unless review file order is intentionally model-visible and must invalidate the cache.
• Touches a guarded path — held for manual review — A maintainer must review and merge this change.

Signal	Result	Evidence
Code review	✅ No blockers	1 reviewer
Linked issue	✅ No-issue rationale	PR body explains why no issue is linked.
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	✅ 20/20	Low review scope from cached public metadata (size label size:XL; no linked issue context).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 67 registered-repo PR(s), 57 merged, 589 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 67 PR(s), 589 issue(s).
Gate result	⚠️ Not blocking	Advisory; not blocking this PR.

Review context

• Author: JSONbored
• Role context: owner (maintainer lane)
• Public audience mode: oss maintainer
• Lane context: Repository registration is not available in the local Gittensory cache.
• Public profile languages: not available
• Official Gittensor activity: 67 PR(s), 589 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Treat this as maintainer-lane context rather than normal contributor-lane activity.
• Triage stale or unlinked PRs.
• No action.
• Link the issue being solved, or explicitly explain why this is a no-issue PR.

Signal definitions

• Related work = same linked issue, overlapping active PRs, or title/path similarity.
• Change scope = cached public metadata such as size labels, draft state, and review-burden hints.
• Validation posture = whether the PR provides enough public validation/test evidence for maintainer review.
• Contributor workload = public contributor activity and cleanup pressure, not a repo-wide quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

---

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by Gittensory, a quiet PR intelligence layer for OSS maintainers.

• Re-run Gittensory review

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Preview URL	Updated (UTC)
✅ Deployment successful! View logs	gittensory-ui	ede5fa1	Commit Preview URL Branch Preview URL	Jul 01 2026, 07:13 PM