Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →- You need to turn a git diff into a reviewer-friendly pull request description.
.claude/skills/comprehensive-review-pr-enhance/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 11 |
| gemini-3.1-pro-preview | 100% | 1 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-18 | ✗→✓ | ▲ Improved | — | — |
| case-14 | ✗→✓ | ▲ Improved | — | — |
| case-17 | ✗→✓ | ▲ Improved | — | — |
| case-19 | ✗→✓ | ▲ Improved | — | — |
git diff <base>...HEAD --stat to identify changed files and scopemarkdown## Summary <!-- one-paragraph executive summary: what changed and why --> ## Changes | Category | Files | Key change | |----------|-------|------------| | source | `src/auth.ts` | added OAuth2 PKCE flow | | test | `tests/auth.test.ts` | covers token refresh edge case | | config | `.env.example` | new `OAUTH_CLIENT_ID` var | ## Why <!-- link to issue/ticket + one sentence on motivation --> ## Testing - [ ] unit tests pass (`npm test`) - [ ] manual smoke test on staging - [ ] no coverage regression ## Risks & Rollback - **Breaking?** yes / no - **Rollback**: revert this commit; no migration needed - **Risk level**: low / medium / high — because ___
Add checklist sections only when the matching file category appears in the diff:
| File category | Checklist items | |---------------|----------------| | source | no debug statements, functions <50 lines, descriptive names, error handling | | test | meaningful assertions, edge cases, no flaky tests, AAA pattern | | config | no hardcoded secrets, env vars documented, backwards compatible | | docs | accurate, examples included, changelog updated | | security-sensitive (auth, crypto, token, password in path) | input validation, no secrets in logs, authz correct |
When diff exceeds 20 files or 1000 lines, suggest splitting by feature area:
git checkout -b feature/part-1
git cherry-pick <commits-for-part-1>resources/implementation-playbook.md — Python helpers for automated PR analysis, coverage reports, and risk scoring| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +50 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.