Code Review Skill

A skill that teaches the AI how to perform thorough code reviews.

Instructions

When asked to review code, follow this systematic approach:

1. First Pass - Overview

• •Understand the purpose of the code
• •Check file/module organization
• •Identify the main logic flow

2. Code Quality Checks

• •Naming: Variables, functions, and classes should have clear, descriptive names
• •DRY: Look for duplicated code that could be refactored
• •SOLID: Check adherence to SOLID principles where applicable
• •Complexity: Flag overly complex functions (cyclomatic complexity)

3. Security Review

• •Check for SQL injection vulnerabilities
• •Look for XSS vulnerabilities in web code
• •Identify hardcoded secrets or credentials
• •Check input validation and sanitization

4. Performance

• •Identify N+1 query problems
• •Look for unnecessary loops or iterations
• •Check for memory leaks or unbounded growth
• •Review async/await usage

5. Error Handling

• •Ensure errors are properly caught and handled
• •Check for bare except/catch blocks
• •Verify error messages are helpful but not leaky

6. Testing

• •Check test coverage
• •Identify untested edge cases
• •Review test quality and assertions

Output Format

Provide your review in this format:

## Summary
[Brief overall assessment]

## Critical Issues 🔴
[Must-fix problems]

## Warnings 🟡
[Should-fix problems]

## Suggestions 🟢
[Nice-to-have improvements]

## Good Practices ✅
[Things done well - positive feedback]

Example Usage

User: "Review this Python code for me" Assistant: Uses this skill's methodology to provide structured review