Rechercher

PythonReactSEOPDF
GitHub Skill

code-review

Run a comprehensive code review of local changes, analyzing staged and unstaged changes to provide insights, identify issues, and suggest alternative approaches. Use when reviewing code before committing or creating pull requests.

kvet☆ 13 Étoiles du dépôt0 Forks du dépôt

Installation

Suivez les instructions du dépôt pour votre agent. Consultez les fichiers et dépendances avant l’installation.

Voir la source ↗Guides

Instructions

Local Code Review

Run a comprehensive, critical code review of local changes. This skill analyzes staged and unstaged changes to provide insights, identify issues, and suggest alternative approaches.

Instructions

When this skill is invoked:

  1. First, read the detailed agent instructions from .claude/agents/code-review/instructions.md
  2. Gather the current changes using git diff
  3. Analyze the changes following the review framework
  4. Provide a structured report with findings and alternatives

Usage

/code-review              # Review all uncommitted changes
/code-review --staged     # Review only staged changes
/code-review <commit>     # Review changes in a specific commit
/code-review <base>..<head>  # Review changes between commits

Review Process

Step 1: Gather Changes

Based on arguments:

  • No args: git diff HEAD (all uncommitted changes)
  • --staged: git diff --staged
  • <commit>: git show <commit>
  • <base>..<head>: git diff <base>..<head>
Step 2: Understand Context

For each modified file:

  • Read the full file to understand surrounding context
  • Identify the module/component's purpose
  • Note existing patterns and conventions
Step 3: Analyze Changes

Apply the review framework from .claude/agents/code-review/instructions.md:

  • Correctness analysis
  • Design evaluation
  • Security review
  • Performance considerations
  • Maintainability assessment
Step 4: Generate Alternatives

For non-trivial changes, brainstorm alternative approaches:

  • Different architectural patterns
  • Alternative algorithms or data structures
  • Trade-off analysis for each approach
Step 5: Report

Provide a structured report with:

  1. Executive summary
  2. Critical issues (must fix)
  3. Concerns (should consider)
  4. Suggestions (nice to have)
  5. Alternative approaches with trade-offs
  6. Questions for the author

Output Format

# Code Review: [brief description of changes]

## Summary

[2-3 sentence overview of what the changes do and overall assessment]

## Critical Issues

[Issues that would cause bugs, security vulnerabilities, or data loss]

## Concerns

[Design issues, edge cases, potential problems]

## Suggestions

[Improvements, style, minor issues]

## Alternative Approaches

### Alternative 1: [Name]

**Approach:** [Description]
**Trade-offs:**

- Pro: ...
- Con: ...
  **When to prefer:** [Conditions where this is better]

### Alternative 2: [Name]

...

## Questions

[Clarifying questions for the author about intent or requirements]

Severity Definitions

  • CRITICAL: Bugs, security issues, data loss risks, breaking changes
  • CONCERN: Design flaws, missing edge cases, unclear intent, potential issues
  • SUGGESTION: Style, readability, minor optimizations, documentation

Exemples

Consultez les exemples du dépôt source.

Voir la source ↗

Historique source

Date d’indexation
2026-01-20
Mise à jour de l’index
2026-01-26
Branch
main

Les dates concernent ce répertoire. Consultez le dépôt pour l’historique du code.

GitHub ↗

chat Comments (0)

chat_bubble_outline

No comments yet. Be the first to share your thoughts!