Performs multi-dimensional AI code reviews on local changes or pull requests using parallel specialized agents.
The Local Code Reviewer skill automates the feedback loop by analyzing uncommitted local changes or specific Pull Requests across five critical domains: code quality, performance, test coverage, documentation, and security. By deploying specialized sub-agents in parallel, it provides deep-dive analysis that goes beyond simple diff checks, cross-referencing related modules and project history. It generates structured Markdown reports and can even apply automated fixes, making it an essential tool for developers who want to maintain high standards before pushing code to a repository.
主な機能
01Direct integration with VSCode for viewing reports and applying fixes
02Parallel multi-agent analysis for quality, performance, testing, docs, and security
031 GitHub stars
04Automatic generation of detailed Markdown reports with actionable suggestions
05Contextual awareness through CLAUDE.md and project-specific rule integration
06Support for local uncommitted changes, specific branches, or PR numbers
ユースケース
01Deep performance and security screening for complex refactoring tasks
02Automated internal audits of pull requests before they undergo human review
03Pre-commit code validation to ensure local work meets security and quality standards
