devs:code-review

Code Review

Comprehensive code review for quality, performance, and architecture across all languages.

Quick Start

Automated Review

Run complete analysis:

./scripts/review_code.sh /path/to/code

This runs:

• Language-specific linters (ESLint, Pylint, Clippy)
• Complexity analysis
• Code smell detection
• Report generation

Manual Review

Use comprehensive checklist:

• See review-checklist.md for complete checklist

Individual Analyzers

Complexity analysis:

./scripts/analyze_complexity.py /path/to/code

Code smells:

./scripts/detect_code_smells.py /path/to/code

When to Use

Trigger this skill when:

• User says "review this code" or similar
• After significant code implementation
• Before committing/creating PRs
• Analyzing unfamiliar codebases
• Investigating quality issues
• Optimizing performance
• Refactoring

What This Skill Provides

Automated Analysis Scripts

review_code.sh - Complete review orchestrator

• Detects project type (JS/TS/Python/Rust)
• Runs appropriate linters
• Analyzes complexity
• Detects code smells
• Generates comprehensive report

analyze_complexity.py - Complexity metrics

• Cyclomatic complexity
• Function length
• Nesting depth
• Identifies problematic functions

detect_code_smells.py - Code smell detection

• Long parameter lists
• Magic numbers
• Duplicated code
• Deep nesting
• God classes

generate_review_report.py - Report generation

• Aggregates all findings
• Prioritizes issues
• Provides recommendations

Manual Review Guidance

review-checklist.md - Complete review checklist

• Code quality checks
• Performance considerations
• Architecture review
• Security checklist
• Testing coverage

code-smells.md - Code smell catalog

• Common smells by category
• Detection strategies
• Refactoring solutions

Review Focus Areas

1. Code Quality

• Naming and readability
• DRY (Don't Repeat Yourself)
• YAGNI (You Aren't Gonna Need It)
• Consistent style
• Proper comments

2. Performance

• Algorithmic complexity
• Unnecessary loops/nesting
• Database query optimization
• Memory management
• Caching opportunities

3. Architecture

• Modularity and organization
• SOLID principles
• Coupling and cohesion
• Dependency management
• Design patterns

4. Code Smells

• Bloaters (long methods, large classes)
• Object-orientation abusers
• Change preventers
• Dispensables (dead code, duplication)
• Couplers (tight coupling)

Common Workflows

Workflow 1: Complete Project Review

# Run automated analysis
./scripts/review_code.sh /path/to/project

# Review output
cat .code-review-output/REVIEW.md

# Manual checklist review
# Consult references/review-checklist.md

# Address findings
# Refactor based on recommendations

Workflow 2: Quick Complexity Check

# Analyze complexity
./scripts/analyze_complexity.py /path/to/code

# Identify complex functions
# Refactor functions with complexity >10

Workflow 3: Pre-Commit Review

1. Run automated review
2. Check for high-severity issues
3. Review manual checklist
4. Ensure tests pass
5. Commit if approved

Workflow 4: Legacy Code Analysis

1. Run full review to understand codebase
2. Identify problematic areas
3. Prioritize refactoring
4. Address incrementally

Understanding Results

Complexity Metrics

Cyclomatic Complexity:

• 1-5: Simple, easy to test
• 6-10: Moderate complexity
• 11-20: High complexity, consider refactoring
• 21+: Very high, definitely refactor

Function Length:

• ≤20 lines: Excellent
• 21-50 lines: Good
• 51-100 lines: Consider splitting
• 100+ lines: Refactor

Nesting Depth:

• 1-2 levels: Good
• 3-4 levels: Acceptable
• 5+ levels: Refactor

Code Smell Severity

High - Address immediately:

• God classes
• Duplicated critical logic
• Deep nesting in hot paths

Medium - Address soon:

• Long parameter lists
• Feature envy
• Duplicated code

Low - Address when convenient:

• Magic numbers
• Minor duplication
• Speculative generality

Language Support

Detected Automatically

JavaScript/TypeScript:

• ESLint integration
• TypeScript compiler checks
• Complexity analysis
• Code smell detection

Python:

• Pylint integration
• Flake8 support
• MyPy type checking
• Complexity analysis

Rust:

• Clippy integration
• Cargo check
• Basic complexity analysis

Additional Languages

For languages not automatically detected:

• Manual checklist still applies
• Use language-specific linters
• Complexity principles are universal

Best Practices

1. Review Early - Don't wait until code is "complete"
2. Focus on Impact - High-severity issues first
3. Be Constructive - Suggest improvements, not just problems
4. Consider Context - Performance vs. readability tradeoffs
5. Automate - Use tools before manual review
6. Test Changes - Ensure refactoring doesn't break functionality
7. Document Decisions - Explain WHY, not just WHAT
8. Iterate - Re-review after changes

Integration

CI/CD Integration

Add to GitHub Actions:

- name: Code Review
  run: |
    ./scripts/review_code.sh .
    # Fail if high-severity issues found

Pre-commit Hook

#!/bin/bash
./scripts/review_code.sh . --quick

Editor Integration

Most linters integrate with VS Code, IntelliJ, etc.

Limitations

• Static Analysis Only - Doesn't catch runtime issues
• Heuristic-Based - Some false positives possible
• Language Coverage - Best for JS/TS/Python/Rust
• No Business Logic - Can't evaluate correctness
• Context Unaware - Doesn't understand project-specific conventions

Getting Help

For specific issues:

• Complexity problems → See complexity thresholds above
• Code smells → See code-smells.md
• General quality → See review-checklist.md

After review:

1. Address high-severity issues
2. Refactor complex functions
3. Remove code smells
4. Re-run review to verify

Example Output

# Code Review Report

## Executive Summary
- Total Functions Analyzed: 127
- Code Smells Detected: 23
  - High Severity: 3
  - Medium Severity: 15
  - Low Severity: 5

## Complexity Analysis
- Average Complexity: 4.2
- Maximum Complexity: 18
- Functions Needing Attention: 8

## Recommendations
- Refactor 3 highly complex functions
- Address 3 high-severity code smells
- Extract duplicated code in 5 locations

---

Note: This skill provides guidance and automated checks. Final decisions on code quality depend on project context, team conventions, and business requirements.