What problem does it solve? Code reviews are often inconsistent, with reviewers missing security vulnerabilities, anti-patterns, or AI-generated code issues. This Skill provides a structured checklist that standardizes reviews across correctness, security, performance, testing, and documentation. ## Core Features & Use Cases - Comprehensive Review Checklist: Covers correctness, security (including prompt injection and output sanitization for AI code), performance, code quality, testing, and documentation. - AI & LLM Review Patterns: Flags hallucination-prone logic, unsafe prompt engineering, and unvalidated AI outputs in modern codebases. - Anti-Pattern Detection: Identifies magic numbers, deep nesting, long functions, and untyped code with concrete before/after examples. - Standardized Review Comments: Uses a severity-tagged comment convention (blocking, suggestion, nit, question) for clear feedback. - Use Case: Before merging a pull request containing AI-generated API handlers, run this checklist to catch missing input validation, hardcoded secrets, and unsanitized LLM outputs. ## Quick Start Review the changes in this pull request using the code review checklist and flag any security or correctness issues.