ai-review

---

name: ai-review description: Review code changes with focus on AI best practices and security license: MIT compatibility: opencode metadata: audience: developers workflow: code-review

What I do

• Review code changes for AI-specific best practices
• Check for common security issues in AI integrations
• Verify proper error handling and API key management
• Ensure documentation meets AI project standards
• Validate testing approaches for AI components

When to use me

Use this skill when reviewing pull requests or code changes that involve:

• AI model integrations
• API calls to AI services
• Configuration changes for AI tools
• New AI agent implementations

Ask clarifying questions if the scope of changes is unclear or if specific security concerns should be prioritized.

Review checklist

1. Security

   • No hardcoded API keys or secrets
   • Environment variables used properly
   • Input validation for AI responses

2. Error Handling

   • Graceful degradation when AI services fail
   • Proper timeout handling
   • Retry logic where appropriate

3. Documentation

   • Clear usage examples
   • API key requirements documented
   • Configuration options explained

4. Testing

   • Mock AI responses in tests
   • Test error scenarios
   • Validate edge cases