fix(review): short-circuit the retry loop on a deliberate INCOHERENT_DIFF_ASSESSMENT bail

[JSONbored]

Summary

ai_review_inconclusive and ai_review_provider_unparseable_exhausted errors have been recurring (Sentry LOOPOVER-1P/2B/29). One observed case shows the model returning a clean, well-formed response with "assessment": "Cannot review — the diff appears out of sync with the PR head.", "blockers": [], "nits": [], "suggestions": [] — the model's own deliberate INCOHERENT_DIFF_ASSESSMENT bail (its prompt explicitly instructs it to emit this exact text rather than rubber-stamp a diff it can't map to the PR). parseModelReview correctly treats this the same as a genuine parse failure (null, since neither yields a usable review) — but the retry loop doesn't distinguish the two, so it burns the full 3-attempt-per-model retry budget (6 attempts across primary + fallback) re-asking the same model the same question before giving up.

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 (e.g. Closes #123) — a linked open issue is required for every contributor PR.

Closes #7518

Validation

• git diff --check
• npm run typecheck
• npm run test:coverage locally — 3 new tests in test/unit/ai-review.test.ts: the retry loop stops after one bail (fallback still gets its own full budget), both models bailing exhausts in 2 attempts total (not 6), and a dedicated unit test for the new isIncoherentDiffBail helper covering the JSON-shape edge cases (no JSON object, malformed JSON hitting the parse catch, a non-string assessment, a look-alike-but-different assessment). New code is 100% covered (verified via targeted per-line coverage analysis); full ai-review.test.ts (204 tests) passes.
• npm run branding-drift:check
• New or changed behavior has unit/integration tests for new branches, fallback paths, and sanitizer boundaries

If any required check was skipped, explain why:

• Same pre-existing, local-environment-only failures noted in earlier PRs this session (Node version mismatch, 2 unrelated miner-CLI test files) — confirmed unrelated, not present in real CI.

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.)
• API/OpenAPI/MCP behavior is updated and tested where needed. (N/A — internal retry-loop logic only.)
• UI changes use live API data or real empty/error/loading states, not production mock/demo fallbacks. (N/A — no UI changes.)
• Visible UI changes include a UI Evidence section. (N/A.)
• Public docs/changelogs are updated where needed; changelogs are only edited for release-prep PRs. (N/A.)

Notes

• New isIncoherentDiffBail helper is a small, standalone check that mirrors parseModelReview's own JSON extraction rather than changing that function's return contract — keeps the change localized to the one retry loop that actually wastes attempts on this (a second, single-shot call site to parseModelReview has no retry budget to save, so it's untouched).

[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 88.46%. Comparing base (784c9aa) to head (9f619eb).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7520      +/-   ##
==========================================
- Coverage   88.48%   88.46%   -0.02%     
==========================================
  Files         720      720              
  Lines       75718    75725       +7     
  Branches    22540    22543       +3     
==========================================
- Hits        66997    66993       -4     
- Misses       7677     7684       +7     
- Partials     1044     1048       +4     

Flag	Coverage Δ
shard-1	32.93% <0.00%> (-0.12%)	⬇️
shard-2	36.86% <100.00%> (+<0.01%)	⬆️
shard-3	25.64% <0.00%> (-0.02%)	⬇️
shard-4	38.89% <0.00%> (-0.01%)	⬇️
shard-5	34.21% <85.71%> (+<0.01%)	⬆️
shard-6	36.23% <14.28%> (-0.01%)	⬇️

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

Files with missing lines	Coverage Δ
src/services/ai-review.ts	96.79% <100.00%> (+0.03%)	⬆️

... and 2 files with indirect coverage changes