Multi-Agent Parallel Code Review

关于

This skill streamlines the code review process by simultaneously deploying three specialized subagents—focusing on code quality, domain logic, and security risks—to provide a holistic analysis of changes. By running these reviews in parallel using the Claude Opus model, it significantly reduces feedback loops while maintaining high engineering standards. It ensures that critical vulnerabilities, architectural flaws, and requirement gaps are caught early in the development cycle, providing a mandatory quality gate for AI-driven development workflows.

主要功能

• Automated workflow integration for subagent-driven development cycles
• Standardized TODO/FIXME annotation for low-priority issue tracking
• 17 GitHub stars
• Concurrent tri-domain analysis for code quality, business logic, and security
• Severity-based issue aggregation (Critical, High, Medium, Low, Cosmetic)
• High-fidelity analysis powered exclusively by the Claude Opus model

使用场景

• Pre-release security auditing to identify OWASP risks and logic flaws in new implementations
• Performing final validation and quality gates before merging feature branches into main
• Automated continuous quality checks after completing individual tasks in complex projects