code-reviewer

name	code-reviewer
description	Advanced code analysis agent that performs comprehensive code reviews focusing on security, performance, maintainability, and best practices. Generates actionable feedback with severity levels and automated fix suggestions.

Code Reviewer Agent Skill

Overview

Advanced code analysis agent that performs comprehensive code reviews focusing on security, performance, maintainability, and best practices. Generates actionable feedback with severity levels and automated fix suggestions.

Capabilities

• Security Analysis: Vulnerability detection, unsafe patterns, authentication issues
• Performance Review: Bottleneck identification, optimization opportunities, resource usage analysis
• Code Quality: Maintainability assessment, complexity analysis, duplication detection
• Best Practices: Language-specific conventions, framework compliance, testing coverage
• Type Safety: Type checking, null safety, generic usage validation

Tools Required

MCP Tools

• codex_read_file - Source code reading and analysis
• codex_grep - Pattern-based code searching
• codex_codebase_search - Semantic code search across project
• grep - Text pattern matching
• read_file - File content access
• codebase_search - Project-wide search

File System Access

• Read: Full codebase access for analysis
• Write: Limited to ./artifacts, ./code-review-reports, ./review-results

Network Access

• https://docs.rs/* - Rust documentation and crate info
• https://doc.rust-lang.org/* - Official Rust documentation
• https://docs.python.org/* - Python standard library docs
• https://developer.mozilla.org/* - Web standards and APIs
• https://cwe.mitre.org/* - Common Weakness Enumeration database

Shell Commands

• cargo check - Rust compilation checking
• cargo clippy - Rust linting
• npm audit - Node.js security audit
• python -m py_compile - Python syntax checking
• rustfmt --check - Code formatting validation

Usage Examples

Basic Code Review

codex $code-reviewer "Review the authentication module for security issues"

Comprehensive Analysis

codex $code-reviewer "Perform full code review: security, performance, and maintainability analysis of the entire codebase"

Language-Specific Review

codex $code-reviewer "Review Rust code for unsafe patterns, lifetime issues, and performance bottlenecks"
codex $code-reviewer "Review Python code for type hints, error handling, and best practices"

CI/CD Integration

codex $code-reviewer "Analyze pull request changes for breaking changes and quality regressions"

Output Format

Review Report Structure

## Code Review Summary

### Severity Breakdown
- 🔴 Critical: X issues
- 🟠 High: X issues
- 🟡 Medium: X issues
- 🟢 Low: X issues

### Categories
#### Security Vulnerabilities
- [ISSUE-001] SQL Injection vulnerability in user input handling
  - Location: `src/auth/login.rs:45`
  - Severity: Critical
  - Recommendation: Use parameterized queries

#### Performance Issues
- [ISSUE-002] Inefficient database query in hot path
  - Location: `src/api/user_handler.rs:123`
  - Severity: High
  - Impact: N+1 query problem
  - Fix: Implement batch loading

#### Code Quality
- [ISSUE-003] Missing error handling for file operations
  - Location: `src/utils/file_ops.rs:67`
  - Severity: Medium
  - Recommendation: Add proper error propagation

### Automated Fixes Available
- [FIX-001] Add input sanitization
- [FIX-002] Implement query optimization
- [FIX-003] Add error handling wrapper

JSON Output (for CI/CD)

{
  "summary": {
    "total_issues": 15,
    "critical": 2,
    "high": 4,
    "medium": 6,
    "low": 3,
    "auto_fixable": 8
  },
  "issues": [
    {
      "id": "ISSUE-001",
      "category": "security",
      "severity": "critical",
      "file": "src/auth/login.rs",
      "line": 45,
      "description": "SQL Injection vulnerability",
      "recommendation": "Use parameterized queries",
      "auto_fix_available": true
    }
  ],
  "metrics": {
    "complexity_score": 4.2,
    "test_coverage_estimate": 85,
    "maintainability_index": 72
  }
}

Progressive Disclosure

Level 1: Quick Scan

Basic syntax, security vulnerabilities, critical issues only.

Level 2: Standard Review

Full analysis including performance, maintainability, best practices.

Level 3: Deep Analysis

Advanced patterns, architectural issues, long-term maintainability.

Configuration

Review Rules Customization

{
  "rules": {
    "security": {
      "enabled": true,
      "severity_threshold": "medium"
    },
    "performance": {
      "enabled": true,
      "complexity_threshold": 10
    },
    "maintainability": {
      "enabled": true,
      "min_test_coverage": 80
    }
  },
  "languages": ["rust", "python", "typescript"],
  "output_format": "markdown"
}

Integration Points

GitHub Actions

- name: Code Review
  uses: openai/codex-code-review@v1
  with:
    paths: "src/"
    fail_on: "critical"
    output_format: "sarif"

VS Code Extension

{
  "contributes": {
    "commands": [
      {
        "command": "codex.codeReview",
        "title": "Run Code Review"
      }
    ]
  }
}

References

• OpenAI Codex Documentation
• Agent Skills Guide
• OWASP Code Review Guide
• Microsoft SDL
• Rust Security Guidelines

---

Installation: $ codex $skill-install https://github.com/zapabob/codex-code-reviewer-skill Version: 2.1.0 Compatibility: Codex v2.9.0+