What is the Multi Review skill?

Multi Review is a Claude Code skill that orchestrates several specialized AI agents to perform a thorough, multi-layered review of your code changes or pull requests.

Is Multi Review suitable for large pull requests?

Yes, it is specifically optimized for large, high-risk, or cross-cutting changes that require more depth and specialized attention than a single AI pass can provide.

Can I use Multi Review with GitHub PRs?

Yes, it can fetch metadata from GitHub PRs to inform the review process and can help format comments to be posted directly back to the pull request.

How does it select which review agents to use?

It performs a discovery phase to analyze your code diff's languages and complexity, then automatically searches for and selects the most relevant installed review plugins.

How does it handle false positives in reviews?

The skill includes a dedicated validation phase where each finding is assessed for accuracy and severity, ensuring the final summary is high-signal and actionable.

Multi Review

Name: Multi Review
Author: gchiam

bygchiam

0•

协作工具

Orchestrates multiple specialized AI agents to perform comprehensive, multi-pass code reviews on pull requests and complex branch diffs.

Multi Review is a sophisticated orchestration skill designed for high-stakes code reviews where a single-pass analysis is insufficient. It automatically discovers available specialized review agents, triages your code changes to select the most relevant reviewers (such as security experts or logic validators), and executes them in parallel. By incorporating a dedicated validation phase to filter false positives and an aggregation phase to de-duplicate findings, it provides a unified, prioritized report of critical issues and architectural improvements, making it an essential tool for large-scale PRs and cross-cutting repository changes.

主要功能

01Parallel execution with real-time polling and background task management.

02Multi-agent orchestration for specialized coverage across security, logic, and style.

03Integrated validation phase to minimize noise and filter out false positives.

04Automated discovery and triage of review agents based on the specific diff profile.

05Comprehensive aggregate summaries categorized by severity and impact.

060 GitHub stars

使用场景

01Standardizing code quality and logic validation across multi-language repositories.

02Reviewing large, complex pull requests with significant architectural impacts.

03Conducting high-risk security audits on sensitive code paths and new interfaces.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gchiam/claude-code-plugins multi-review

For use in Claude.ai and ChatGPT

Download Skill