Notre avis
Extrait des conseils d'implémentation et évalue les risques d'une Pull Request avant de reproduire la correction.
Points forts
- Structure claire pour analyser rapidement une PR
- Inclut des critères de qualité, tests, documentation et breaking changes
- Évalue la complexité et les signaux d'alerte
- Fournit une base pour créer un plan de test
Limites
- Ne remplace pas une revue manuelle approfondie
- La note de relation PR/issue peut être subjective
- Pas adapté aux PRs externes (merge gate)
Utilisez cette checklist quand vous voulez extraire des enseignements d'une PR pour votre propre implémentation ou évaluer si une PR est une bonne référence.
Ne l'utilisez pas comme un processus de validation obligatoire pour fusionner des PRs externes.
Analyse de sécurité
SûrThis skill is a static checklist and template for PR analysis and code review. It does not instruct the agent to execute any commands, access external resources, or handle sensitive data. There are no destructive or exfiltration risks.
Aucun point d'attention détecté
Exemples
Review this PR using the quality checklist and extract all implementation guidance, risk notes, and a test plan.Analyze the attached PR for risk level, complexity signals, and breaking changes. Summarize what to be careful about before implementing a similar fix.Based on the changes in this PR, create a test plan covering happy path, edge cases, and error handling using the testing criteria from the quality checklist.Quality Checklist
At a glance
- Purpose: Extract implementation guidance from PRs and assess risk before you implement the fix yourself.
- Use when: You are deciding whether a PR is a good “reference implementation,” or you’re planning tests/edge cases.
- Output: Risk notes + test plan + any constraints to preserve.
- Note: This is not a merge gate for external PRs.
Use this checklist to extract insights and assess risk. It is not a merge gate for external PRs.
PR Analysis Checklist
Use this when extracting insights from PRs to inform your own implementation.
Automated Signals
Check CI/CD results (indicates if their approach works):
- [ ] Tests pass
- [ ] Linting passes
- [ ] Build succeeds
- [ ] Coverage maintained or improved
Code Quality
- [ ] Logic is correct (actually solves the stated problem)
- [ ] Edge cases handled appropriately
- [ ] Error handling is adequate (not silent failures)
- [ ] No obvious performance issues
- [ ] No security vulnerabilities introduced
Style and Patterns
- [ ] Follows existing code patterns in the repo
- [ ] Consistent naming conventions
- [ ] Appropriate abstraction level (not over/under-engineered)
- [ ] Comments where logic is non-obvious
Testing
- [ ] New functionality has tests
- [ ] Tests are meaningful (not just coverage padding)
- [ ] Tests cover both happy path and error cases
- [ ] Tests are maintainable and readable
Documentation
- [ ] Public API changes documented
- [ ] README updated if user-facing behavior changed
- [ ] Changelog entry if needed
- [ ] Code comments for complex logic
Relationship Quality
- [ ] PR intent matches linked issue intent
- [ ] No mismatch between reported symptoms and implemented fix
- [ ] Relationship quality noted (strong/medium/weak)
Breaking Changes
If breaking changes present:
- [ ] Clearly documented in PR description
- [ ] Migration path provided
- [ ] Version bump appropriate (semver)
Risk Assessment
File Risk Levels
| Location | Risk | Reason | |----------|------|--------| | Core/critical paths | HIGH | Wide impact, careful review | | User-facing features | MEDIUM | Visible bugs, moderate impact | | Tests | LOW | Isolated, easy to fix | | Documentation | LOW | No runtime impact | | Configuration | VARIES | Depends on what's configured |
Change Size
| Size | Files | Lines | Review Approach | |------|-------|-------|-----------------| | Small | 1-3 | <50 | Quick review | | Medium | 4-10 | 50-200 | Careful review | | Large | 10+ | 200+ | Consider splitting |
Complexity Signals
Red flags requiring extra scrutiny:
- Refactoring + new features in same PR
- Changes to error handling or retry logic
- Concurrency or async changes
- Security-sensitive areas (auth, input validation)
- Database schema changes
- API contract changes
Issue Quality Assessment
Well-Formed Bug Report
- [ ] Clear title describing the problem
- [ ] Steps to reproduce
- [ ] Expected behavior
- [ ] Actual behavior
- [ ] Environment details (OS, version, etc.)
- [ ] Error messages/stack traces if applicable
Well-Formed Feature Request
- [ ] Clear description of desired functionality
- [ ] Use case / motivation explained
- [ ] Consideration of alternatives
- [ ] Scope is reasonable
Actionability Score
| Score | Criteria | |-------|----------| | High | All needed info present, can start work immediately | | Medium | Most info present, minor clarification needed | | Low | Missing critical info, needs follow-up | | None | Too vague to act on |
Agent Brief Quality
When creating agent briefs, ensure:
- [ ] Clear problem statement (what to solve)
- [ ] Specific acceptance criteria (how to verify)
- [ ] Relevant context links (issues, PRs, files)
- [ ] Constraints noted (don't change X, maintain compatibility with Y)
- [ ] Scope bounded (what's NOT in scope)
Agent Brief Template
## Task: [Title]
**Intent:** [What problem this solves]
**Context:**
- Related issue: [ISSUE:XX]
- Key files: `src/path/file.ts`
**Acceptance Criteria:**
- [ ] [Specific, testable criterion]
- [ ] [Another criterion]
**Constraints:**
- Do not change [X]
- Must maintain backwards compatibility with [Y]
**Out of Scope:**
- [Related but excluded work]
Repo Health Checklist
Use during opportunity scans:
- [ ] README has a clear quickstart and examples
- [ ] Docs match current behavior
- [ ] CI is reliable and fast enough
- [ ] Tests cover critical paths
- [ ] Releases are documented and easy to follow
Expert Next.js App Router
Developpement
Un skill qui transforme Claude en expert Next.js App Router.
Générateur de README
Developpement
Crée des README.md professionnels et complets pour vos projets.
Rédacteur de Documentation API
Developpement
Génère de la documentation API complète au format OpenAPI/Swagger.