feat(commands): wire @gittensory review and resume PR-comment commands

[JSONbored]

Summary

• Wires the last two handlers in the @gittensory PR-comment command surface: maybeProcessReviewCommand and maybeProcessResumeCommand, both mirroring the existing classify → authorize → dispatch shape used by pause/resolve/explain.
• review (alias re-review) dispatches to the existing reReviewStoredPullRequest path with force: true, so a maintainer explicitly asking for a fresh pass bypasses the AI-review cache/dedup instead of getting a stale verdict. It never touches the Gate check-run, AgentActionMode, or the one-shot disposition — whatever the re-review's own gate evaluation produces is exactly what a scheduled sweep pass would produce.
• resume fixes a real bug in hasAutoreviewPausedMarker: it previously only checked whether an autoreview_paused row had ever been recorded, so a resume command could authorize and post a confirmation comment but silently fail to actually resume anything — the next re-review pass would still see the stale pause as active forever. It now reads the most recent of {autoreview_paused, autoreview_resumed} for the target (order by created_at desc, rowid desc), with the rowid tiebreaker covering same-millisecond writes from back-to-back commands in a test/rapid-succession scenario.

Scope

Closes #2163, Closes #2165, Closes #1960 (the 10th and final sub-issue completed under the parent — command registration, authorization policy, pause/resolve/explain/configuration/help handlers, and the did-you-mean fallback all shipped in prior PRs).

Validation

• npx vitest run test/unit/queue.test.ts — 714/714 passing, including ~10 new tests covering authorized dispatch, re-review alias resolution, unauthorized denial, missing-PR skip, and bot-author skip for review; and authorized resume superseding an earlier pause, full pause→resume→pause-again ordering, unauthorized denial, missing-PR skip, and bot-author skip for resume.
• Deleted one pre-existing test (#2160) whose entire premise — "a catalog-registered verb with no dispatch handler exists" — is no longer true now that all 7 registered verbs have handlers; not a coverage dodge, the scenario it guarded against can no longer occur.
• npx tsc --noEmit clean.
• npm run docs:drift-check clean (no gate-mode/command-catalog field added; review/resume were already registered verbs pre-dating this PR).
• git diff --check clean.
• Rebased onto latest origin/main, no conflicts.

Safety

• Both commands are strictly auto-review-scoped, per PR command surface — @gittensory review / pause / resume / resolve / help / configuration / explain #1960's hard constraint: neither can flip the Gate to advisory or bypass the one-shot disposition.
• hasAutoreviewPausedMarker still fails open (catch returns false) so a corrupt/unreachable audit ledger can't wedge review processing.

Notes

No generated-artifact regeneration needed (no API/schema, wrangler binding/var, src/selfhost/** env read, or DB schema changes in this PR).

[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 93.73%. Comparing base (4664ad2) to head (a93956f).
⚠️ Report is 3 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4175   +/-   ##
=======================================
  Coverage   93.72%   93.73%           
=======================================
  Files         387      387           
  Lines       36320    36371   +51     
  Branches    13306    13316   +10     
=======================================
+ Hits        34042    34092   +50     
  Misses       1621     1621           
- Partials      657      658    +1     

Files with missing lines	Coverage Δ
src/queue/processors.ts	95.30% <100.00%> (+0.04%)	⬆️

🚀 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: 2026-07-08 09:53:33 UTC}

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

⏸️ Suggested Action - Manual Review

• Touches a guarded path — held for manual review: This PR changes guardrail-protected path(s): src/queue/processors.ts (matched src/queue/**).

Review summary
This PR wires the final two @​gittensory PR-comment commands: `review`/`re-review` (dispatches reReviewStoredPullRequest with force:true, bypassing cache without touching Gate/disposition) and `resume` (which also fixes a real latent bug in hasAutoreviewPausedMarker — the old query only checked whether a pause row had ever existed, so resume could authorize and post a confirmation while never actually clearing the paused state). The new query now reads the most-recent of {autoreview_paused, autoreview_resumed} ordered by created_at desc, rowid desc, correctly letting resume supersede an earlier pause while still allowing a later pause to re-pause. Both handlers mirror the existing classify → authorize → dispatch shape used by pause/resolve/explain, and the new tests exercise the real paths (a live PR resync fetch proves reReviewStoredPullRequest actually ran, and pause→resume→pause-again ordering is asserted against the real query) rather than fabricating unreachable states.

Nits — 7 non-blocking

• The classify → authorize → dispatch boilerplate is now duplicated four times (pause/resolve/review/resume) with only the audit event names and confirmation text differing — consider extracting a shared `dispatchPrCommand` helper now that the pattern is fully established across all four handlers.
• processors.ts:11094+ uses non-null assertion `payload.issue!` in both new handlers; this mirrors the existing pause/resolve pattern so it's consistent, but worth confirming GitHub always sends `issue` on `issue_comment` events for PR threads (it does for PRs, just flagging the assumption).
• test/unit/queue.test.ts's `isCurrentlyPaused` test helper duplicates the exact SQL from `hasAutoreviewPausedMarker` rather than importing it — acceptable given the file's existing convention of raw audit_events queries in tests, but a shared test utility would reduce drift risk if the query ever changes again.
• The review-command confirmation comment text ('Re-running auto-review for this PR...') and resume's ('Auto-review is resumed for this PR...') are hand-written strings in the handler bodies; consider centralizing PR-comment templates the way other command confirmations likely already do, for consistency if wording needs a global update later.
• Extract the common classify/authorize/audit-skip scaffolding shared by maybeProcessReviewCommand, maybeProcessResumeCommand, maybeProcessPauseCommand, and maybeProcessResolveCommand into a single higher-order dispatcher to cut ~30 duplicated lines per handler.
• PR author also opened the linked issue — Link an issue that was opened by a different contributor, or provide a rationale for why this self-authored issue represents genuine discovery work.
• 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	✅ Linked	#2163, #2165, #1960
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 (3 linked issues).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 52 registered-repo PR(s), 43 merged, 486 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	JSONbored; Gittensor profile; 52 PR(s), 486 issue(s).
Gate result	⚠️ Not blocking	Advisory; not blocking this PR.

Linked issue satisfaction

Addressed
The diff adds maybeProcessReviewCommand mirroring the existing classify→authorize→dispatch shape, registers it in the comment-dispatch chain, calls the existing reReviewStoredPullRequest path (with force:true) rather than reimplementing review, records review_command_completed/denied/skipped audit and product-usage events, and includes unit tests for authorized dispatch, the re-review alias, unaut

Review context

• Author: JSONbored
• Role context: owner (maintainer lane)
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: Python, TypeScript, JavaScript, Ruby, Go, Kotlin, MDX, Shell
• Official Gittensor activity: 52 PR(s), 486 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Treat this as maintainer-lane context rather than normal contributor-lane activity.

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