Agent skill
code-reviewer
6-aspect structured code review. Checks security, types, error handling, tests, quality, simplification.
Install this agent skill to your Project
npx add-skill https://github.com/nguyenthienthanh/aura-frog/tree/main/aura-frog/skills/code-reviewer
SKILL.md
Aura Frog Code Reviewer โ 6-Aspect Analysis
Priority: HIGH โ Use before merging code
When to Use
- After implementation, before merge
- During Phase 4 (Refactor + Review)
- When explicitly requested
Review Process
Step 1: Get Changed Files
git diff --name-only main...HEAD
# Or: files modified in current workflow
Step 2: Run 6-Aspect Review
MANDATORY: All 6 aspects must be covered. Do not skip any.
Aspect 1: ๐ Security
- Hardcoded secrets (API keys, passwords, tokens)
- SQL injection, XSS, command injection vectors
- Auth/authz gaps (missing middleware, privilege escalation)
- CSRF, CORS misconfigurations
- Insecure crypto (MD5, SHA1, Math.random for tokens)
Aspect 2: ๐ท๏ธ Type Safety
- Missing type annotations on public functions
anytype usage (suggest specific types)- Inconsistent return types
- Null/undefined handling gaps
- Generic types that could be narrower
Aspect 3: โ ๏ธ Error Handling
- Unhandled promise rejections
- Empty catch blocks without justification
- Missing error boundaries (React) / error middleware (Express)
- Silent failures (errors swallowed without logging)
- Missing retry logic on external calls
Aspect 4: ๐งช Test Gaps
- Untested critical paths
- Missing edge case tests
- Test quality (testing behavior vs implementation)
- Mock quality (over-mocking, missing integration tests)
- Gaps on modified files
Aspect 5: ๐ Code Quality
- KISS violations (over-engineering, premature abstraction)
- DRY violations (duplicated logic)
- Naming clarity (functions, variables, files)
- Single Responsibility violations
- Dead code, unused imports
Aspect 6: โป๏ธ Simplification Opportunities
- Complex conditionals that could be simplified
- Deep nesting that could be flattened (early returns)
- Long functions that should be split
- Verbose patterns with simpler alternatives
- Redundant null checks or type guards
Step 3: Generate Report
review[6]{aspect,icon,status,findings}:
Security,๐,โ
|โ ๏ธ|โ,{count} findings
Types,๐ท๏ธ,โ
|โ ๏ธ|โ,{count} findings
Errors,โ ๏ธ,โ
|โ ๏ธ|โ,{count} findings
Tests,๐งช,โ
|โ ๏ธ|โ,{count} findings
Quality,๐,โ
|โ ๏ธ|โ,{count} findings
Simplify,โป๏ธ,โ
|โ ๏ธ|โ,{count} findings
Detail each finding:
[ASPECT] [SEVERITY] file:line โ description
โ Fix: recommendation
Severity: ๐ด CRITICAL (block merge) | ๐ก WARNING (should fix) | ๐ต INFO (nice to have)
Step 4: Decision
- โ APPROVED โ 0 critical, โค3 warnings
- โ ๏ธ APPROVED WITH COMMENTS โ 0 critical, >3 warnings
- โ CHANGES REQUESTED โ Any critical finding
Step 5: Summary Line
Review: ๐โ
๐ท๏ธโ
โ ๏ธโ ๏ธ ๐งชโ
๐โ
โป๏ธโ
โ APPROVED WITH COMMENTS (1 error handling warning)
Critical (Block Merge)
- Hardcoded secrets
- SQL injection / XSS / command injection
- Missing auth on protected routes
- Breaking changes without migration
Remember: Review improves code quality. Be constructive.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
scalable-thinking
Design for scale while keeping implementation simple (KISS).
debugging
Systematic debugging with root cause investigation. NO fixes without understanding cause first.
migration-helper
Guide safe database and code migrations with zero-downtime strategies.
testing-patterns
Unified testing patterns across all frameworks. Provides consistent test structure, naming, and best practices for Jest, Vitest, Pytest, PHPUnit, Go testing, and more.
phase1-lite
Ultra-compact Phase 1 requirements output. HARD CAP: 500 tokens.
stitch-design
Generate UI designs using Google Stitch AI with optimized prompts
Didn't find tool you were looking for?