Is it possible to trigger reviews via GitHub comments?

Yes, the skill includes a webhook handler that recognizes /swarm commands in PR comments to initialize, status, or spawn new agents dynamically.

Can it automatically block a Pull Request from merging?

Yes, when integrated with GitHub Actions, the skill can be configured to request changes or fail status checks if critical security or performance issues are detected.

How does this skill differ from standard linters?

Unlike static linters that check for syntax, this skill uses a swarm of specialized AI agents to understand logic, intent, and architectural design, providing deeper insights into security and performance.

Does this skill require external dependencies?

Yes, this skill requires the GitHub CLI (gh), the ruv-swarm package, and the claude-flow framework to orchestrate agent communication and repository interaction.

Can I customize which agents review my code?

Absolutely. You can configure specific agents (like security or accessibility) based on PR labels, file paths, or project-specific risk profiles defined in your configuration.

GitHub AI Code Review Swarm

Name: GitHub AI Code Review Swarm
Author: EarthmanWeb

byEarthmanWeb

•

Collaboration Tools

Automates comprehensive multi-agent code reviews on GitHub to enhance security, performance, and architectural integrity.

This skill leverages the Claude Flow framework and ruv-swarm to deploy a specialized team of AI agents for exhaustive GitHub Pull Request analysis. It transcends traditional static analysis by coordinating individual agents focused on security vulnerabilities, performance bottlenecks, design patterns, and accessibility standards. By integrating directly with the GitHub CLI, it allows teams to enforce high-quality gates, automate feedback cycles, and manage complex PR workflows through simple terminal commands or PR comments, significantly reducing the manual burden on human reviewers.

Key Features

01Seamless GitHub CLI integration for native developer workflows

022 GitHub stars

03Multi-agent swarm orchestration for parallel, domain-specific code analysis

04Interactive PR comment commands for real-time swarm management and status

05Specialized review agents for security, performance, architecture, and style

06Automated Pull Request management and quality gate enforcement

Use Cases

01Enforcing consistent architectural patterns and coding standards across distributed repositories

02Identifying critical security vulnerabilities and OWASP violations before merging code

03Scaling code review capacity for large engineering teams with automated first-pass analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add earthmanweb/claude-flow-plugin github-code-review

For use in Claude.ai and ChatGPT

Download Skill