Automates multi-agent pull request reviews with impact scoring and intelligent inline GitHub comments.
This skill orchestrates a sophisticated multi-phase code review process using specialized AI agents to analyze security, bugs, code quality, and architectural contracts. It intelligently filters out noise by calculating confidence and impact scores for every finding, ensuring that only meaningful, actionable feedback is posted as inline comments directly to your GitHub pull requests. By integrating project-specific context from files like CLAUDE.md, it provides reviews that are tailored to your unique coding standards and historical git context.
Key Features
01542 GitHub stars
02Automated impact and confidence scoring to eliminate low-value nitpicks
03Direct GitHub integration for posting inline line-specific feedback
04Historical context analysis using git blame and previous PR history
05Context-aware evaluation using project guidelines and documentation
06Parallel multi-agent analysis for security, bugs, quality, and API contracts
Use Cases
01Enforcing project-specific architectural standards and style guides
02Reducing reviewer fatigue by filtering out noise and focusing on high-impact issues
03Automated first-pass code reviews to catch bugs and security flaws before human review
