Does this skill work on the entire project or just single files?

It is designed for comprehensive 'full project' reviews, though it can be triggered for specific sections of code depending on your prompt.

How does it handle conflicting feedback from different AIs?

It uses fuzzy matching to group similar issues and categorizes them into Consensus (all agree), Majority (most agree), and Unique findings, helping you prioritize high-confidence items.

Which AI models are supported by this skill?

The skill currently coordinates reviews between Claude, Gemini, and Codex, allowing for a broad spectrum of AI perspectives.

Do I need all three AI CLIs installed to use this skill?

No, the skill will automatically detect which CLIs are available on your system, report missing ones with installation instructions, and proceed with the tools you have ready.

Multi-AI Code Review

Name: Multi-AI Code Review
Author: ThrownLemon

byThrownLemon

•

Collaboration Tools

Performs comprehensive code reviews by aggregating perspectives from Claude, Gemini, and Codex to find consensus and unique vulnerabilities.

The Multi-AI Code Review skill enhances your development workflow by orchestrating a parallel audit across multiple leading AI models including Claude, Gemini, and Codex. By leveraging a specialized review-coordinator, it identifies high-confidence consensus findings, flags issues recognized by the majority, and surfaces unique insights that a single AI might miss. This multi-perspective approach reduces false positives and ensures a more thorough analysis of code quality, security, and performance compared to standard single-model reviews, providing a definitive cross-check of your entire project.

Key Features

01Automated consensus analysis and issue aggregation

02Graceful handling and installation reporting for missing AI CLIs

03Fuzzy matching to identify similar findings across different models

041 GitHub stars

05Parallel execution of Claude, Gemini, and Codex reviews

06Detailed reports categorized by agreement level (Consensus, Majority, Unique)

Use Cases

01Performing high-stakes security audits before production deployment

02Gaining a comprehensive second opinion to validate automated refactoring suggestions

03Comparing model interpretations to resolve ambiguous or complex code patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

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

For use in Claude.ai and ChatGPT

Download Skill