feat(review): size-cap eligibility guard for auto-review via review.auto_review.max_added_lines / max_files

[bohdansolovie]

Summary

• Add review.auto_review.max_added_lines and max_files to the focus manifest so maintainers can skip AI auto-review on oversized PRs (0 = cap disabled).
• Compute added-line totals via totalAddedLineCount / addedLineCount from review diffs and thread counts through resolveAutoReviewSkipForPullRequest, emitting review skipped (too large) audit events.
• Add unit, wiring, precedence-matrix, and queue integration tests including exact-threshold boundaries (eligible at cap, skip when over).

Fixes #2065

Test plan

• npx tsc -p tsconfig.json --noEmit
• review-diff totalAddedLineCount helper tests
• focus-manifest parse/evaluate/round-trip tests for size caps
• auto-review-wiring tests for resolver + resolveAutoReviewSkipForPullRequest
• auto-review-config-matrix precedence case for too-large skip
• queue.test.ts integration test for max_added_lines skip + audit detail

Made with Cursor

[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 93.35%. Comparing base (525be40) to head (c5a9972).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #3693   +/-   ##
=======================================
  Coverage   93.35%   93.35%           
=======================================
  Files         315      315           
  Lines       32164    32187   +23     
  Branches    11795    11806   +11     
=======================================
+ Hits        30026    30049   +23     
  Misses       1507     1507           
  Partials      631      631           

Files with missing lines	Coverage Δ
src/queue/processors.ts	94.06% <100.00%> (+0.01%)	⬆️
src/review/review-diff.ts	91.66% <100.00%> (+0.75%)	⬆️
src/signals/focus-manifest.ts	99.08% <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.

[loopover-orb]

Warning

🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨🟨

⏸️ Gittensory review result - manual review recommended

_{Review updated: 2026-07-06 04:39:29 UTC}

9 files · 1 AI reviewer · no blockers · readiness 80/100 · CI green · clean

⏸️ Suggested Action - Manual Review

Review summary
This PR adds two new focus-manifest knobs (`max_added_lines`, `max_files`) that let maintainers skip AI auto-review on oversized PRs, threading an added-line/file-count total through `evaluateAutoReviewSkipReason` and `resolveAutoReviewSkipForPullRequest`, with a new `totalAddedLineCount` helper aggregating per-file `addedLineCount`. The wiring is complete end-to-end: manifest parse/normalize/round-trip, the resolver, and the queue processor call site in `maybePublishPrPublicSurface` all pass real `addedLineCount`/`changedFileCount` derived from `getReviewFiles()`, and the boundary condition (skip only when strictly `>` the cap, eligible exactly at the cap) is correctly implemented and tested. The `queue.test.ts` integration test exercises the real `processJob` pipeline with a mocked GitHub fetch returning an actual patch, so the added-lines count is genuinely computed from a real diff rather than fabricated — this is the kind of real-path test this repo wants.

Nits — 5 non-blocking

• The externally-flagged `generic_secret_assignment` hit at test/unit/queue.test.ts:3367 is almost certainly the literal mock token string `fake-installation-token` used in the fetch stub, not a real secret — worth a quick glance to confirm but not blocking.
• `totalAddedLineCount`'s dual `patch`/`payload.patch` shape suggests two different file-record shapes feed this helper (raw GitHub API vs. some wrapped payload) — a one-line comment on why both shapes exist would help the next reader avoid re-deriving it.
• The new size-cap check sits after `skipDocsOnly` and before `baseBranches` in `evaluateAutoReviewSkipReason` — the precedence-matrix test only covers this one ordering slot; worth confirming this is the intended maintainer-facing precedence for the docs/config example if one gets added later.
• Consider documenting the new `max_added_lines`/`max_files` keys in `config/examples/gittensory.full.yml` alongside the other `auto_review` fields for discoverability, since that file is the canonical example maintainers copy from.
• If `getReviewFiles()` can be paginated/truncated for very large PRs, note that the size-cap total is computed from whatever page was fetched — not a blocker here since the fetch logic isn't in this diff, but worth a follow-up sanity check given the whole point of this feature is large PRs.

Signal	Result	Evidence
Code review	✅ No blockers	1 reviewer
Linked issue	✅ Linked	#2065
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 (1 linked issue).
Validation posture	❌ 5/25	Preflight is holding this PR: the review lane is unavailable, so it is not ready for automated review.
Contributor workload	✅ 10/10	Author activity: 331 registered-repo PR(s), 195 merged, 8 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	bohdansolovie; Gittensor profile; 331 PR(s), 8 issue(s).
Gate result	✅ Passing	No configured blocker found.

Review context

• Author: bohdansolovie
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository registration is not available in the local Gittensory cache.
• Public profile languages: Python, C++, JavaScript
• Official Gittensor activity: 331 PR(s), 8 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Await review-lane availability.
• Refresh registry data or choose a registered active repo.

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

[loopover-orb]

Gittensory approves — the gate is satisfied and CI is green.