Skill Metadata

Field	Value
Version	1.0.0
Tier	Quality
Auto-load	When conducting code reviews or quality checks

What It Does

체계적인 코드 리뷰 프로세스와 자동화 가이드를 제공합니다. TRUST 5 원칙 적용, 코드 품질 검증, SOLID 원칙 준수, 보안 이슈 식별, 유지보수성 보장을 다룹니다.

When to Use

코드 리뷰를 수행할 때
리뷰 표준과 가이드라인을 설정할 때
코드 품질 자동화를 구현할 때
팀의 코드 리뷰 문화를 개선할 때

Systematic Code Review with TRUST 5 Principles

Code review is a quality assurance process that ensures code meets standards, follows best practices, and maintains long-term maintainability. This skill provides systematic guidance for conducting thorough, effective code reviews.

Code Review Framework

The TRUST 5 Review Framework

Principle	Review Focus	Key Questions	Tools
T - Test First	Test coverage & quality	Are tests comprehensive? Do they test edge cases?	pytest coverage, jest --coverage
R - Readable	Code clarity & maintainability	Is code self-documenting? Are names meaningful?	linters, code formatters
U - Unified	Consistency & standards	Does it follow team patterns? Is it cohesive?	style guides, architectural patterns
S - Secured	Security & vulnerabilities	Are inputs validated? Are secrets handled properly?	security scanners, static analysis
T - Trackable	Documentation & traceability	Is code linked to requirements? Are changes documented?	@TAG system, git history

High-Freedom: Review Strategy & Philosophy

When to Review vs When to Trust

markdown

## Review Decision Matrix

| Change Type | Review Level | Automation | Focus Areas |
|-------------|--------------|------------|-------------|
| Critical security | 🔴 Mandatory | Full scan | Vulnerabilities, input validation |
| Core architecture | 🟡 Deep review | Partial | Design patterns, scalability |
| Bug fixes | 🟢 Standard | Automated | Root cause, test coverage |
| Documentation | 🟢 Light | Basic | Accuracy, completeness |
| Configuration | 🟢 Automated | Full | Security, best practices |

Review Culture Principles

✅ Psychological Safety: Reviews are about code, not people
✅ Learning Opportunity: Reviews transfer knowledge and standards
✅ Constructive Feedback: Focus on improvement, not criticism
✅ Consistent Standards: Apply same criteria to all code
✅ Efficient Process: Automated checks first, human review for value-add

Review Anti-patterns to Avoid

❌ Nitpicking: Focus on style over substance
❌ Authoritarian: "Do it this way because I said so"
❌ Incomplete: "Looks good" without specific feedback
❌ Delayed: Reviews blocking progress for days
❌ Inconsistent: Different standards for different people

Medium-Freedom: Review Process & Checklists

Structured Review Process

pseudocode

## Review Workflow

1. PRE-REVIEW AUTOMATION (2-5 min)
   a. Run linters and formatters
   b. Execute test suite with coverage
   c. Scan for security vulnerabilities
   d. Check for @TAG compliance
   
2. CODE COMPREHENSION (5-10 min)
   a. Read commit message and PR description
   b. Understand the problem being solved
   c. Identify affected components
   d. Review test changes first
   
3. DETAILED REVIEW (10-20 min)
   a. Apply TRUST 5 framework systematically
   b. Check architectural consistency
   c. Validate error handling
   d. Assess performance implications
   
4. FEEDBACK SYNTHESIS (5 min)
   a. Categorize issues: Must-fix, Should-fix, Nice-to-have
   b. Provide specific, actionable feedback
   c. Explain reasoning behind suggestions
   d. Offer to discuss complex changes

Code Review Checklist Template

markdown

## Code Review Checklist

### 🧪 Test Coverage (T)
- [ ] New features have corresponding tests
- [ ] Test coverage ≥ 85% (or team standard)
- [ ] Edge cases and error conditions tested
- [ ] Integration tests included where appropriate
- [ ] Tests are readable and maintainable

### 📖 Readability (R)
- [ ] Function and variable names are descriptive
- [ ] Complex logic is commented or extracted
- [ ] File length ≤ 300 LOC (or team standard)
- [ ] Function length ≤ 50 LOC (or team standard)
- [ ] No magic numbers or hardcoded values

### 🔗 Unity (U)
- [ ] Follows established team patterns
- [ ] Consistent with existing codebase style
- [ ] Uses shared utilities and libraries
- [ ] Architecture aligns with project structure
- [ ] Imports and dependencies are organized

### 🔒 Security (S)
- [ ] Input validation for all user inputs
- [ ] No hardcoded secrets or credentials
- [ ] Proper error handling without information leakage
- [ ] Authentication and authorization checked
- [ ] SQL injection and XSS protection in place

### 🏷️ Traceability (T)
- [ ] Code changes linked to SPEC or issue
- [ ] @TAG references are correct and complete
- [ ] Commit message is clear and descriptive
- [ ] Documentation updated as needed
- [ ] Breaking changes are documented

Low-Freedom: Automated Review Scripts

Pre-commit Review Automation

bash

#!/bin/bash
# .claude/skills/moai-alfred-code-reviewer/scripts/pre-review-check.sh

set -e

echo "🔍 Running automated code review checks..."

# Test Coverage Check
echo "📊 Checking test coverage..."
python -m pytest --cov=src --cov-fail-under=85 --cov-report=term-missing

# Code Quality Checks
echo "🧹 Running linters..."
python -m ruff check src/ --show-source
python -m mypy src/ --strict

# Security Scanning
echo "🔒 Scanning for security issues..."
python -m bandit -r src/ -f json -o bandit-report.json

echo "✅ All automated checks passed!"

Integration with Development Workflow

Pre-commit Integration

yaml

# .pre-commit-config.yaml
repos:
  - repo: local
    hooks:
      - id: security-review
        name: Security Review
        entry: .claude/skills/moai-alfred-code-reviewer/scripts/security_review.py
        language: script
        args: [src/]
        pass_filenames: false

Review Best Practices Summary

For Reviewers

✅ Start with understanding: Read the PR description first
✅ Automate first: Let tools catch the obvious issues
✅ Focus on value: Spend time on architectural and security concerns
✅ Be specific: Provide exact locations and suggestions
✅ Explain why: Help the author understand the reasoning

For Authors

✅ Self-review first: Run all automated checks before submitting
✅ Write clear descriptions: Explain what and why
✅ Keep PRs small: Large changes are harder to review effectively
✅ Respond promptly: Address feedback in a timely manner
✅ Learn from feedback: Use reviews as learning opportunities

Reference: Code Review Best Practices, TRUST 5 Principles
Version: 1.0.0

Search AI Tools

moai-alfred-code-reviewer

Install this agent skill to your Project

SKILL.md