Which AI models are used in the review process?

The orchestrator coordinates Claude Code for architecture, Codex for logic, Gemini for performance, and Factory.ai Droid for security and maintainability.

Does it support iterative improvements?

Yes, the skill runs multiple cycles, applying fixes and re-reviewing the code until the improvement delta falls below a specific threshold or a max iteration limit is reached.

How does the skill handle conflicting AI suggestions?

A master orchestrator agent uses a priority scoring algorithm that weighs agreement levels, severity, and implementation complexity to resolve contradictions.

What are the prerequisites for using this skill?

You need API keys for the respective AI services and the corresponding CLI tools (Codex, Gemini, Droid) configured within your environment.

Can this skill automatically fix my code?

Yes, high-consensus changes (agreed upon by 3 or more AI models) can be automatically applied, while lower-confidence issues are flagged for manual review.

Multi-AI Code Review Orchestrator

Name: Multi-AI Code Review Orchestrator
Author: dev-jelly

bydev-jelly

0•

Collaboration Tools

Orchestrates multiple AI models to perform comprehensive, cross-validated code reviews with automated improvement cycles.

This skill transforms Claude into a master orchestrator that coordinates multiple AI engines—including Claude Code, Codex, Gemini, and Factory.ai Droid—to analyze code from diverse perspectives such as architecture, logic, performance, and security. It goes beyond simple analysis by aggregating multi-model feedback, resolving conflicting suggestions through a priority scoring system, and iteratively applying improvements until the code meets high-quality benchmarks, ensuring production-ready results with minimal manual intervention.

Key Features

01Comprehensive multi-perspective quality and audit reports

02Multi-model orchestration using Claude, Codex, Gemini, and Droid

03Iterative improvement cycles with automated validation

04Intelligent conflict detection and automated resolution logic

05Priority-based scoring for severity and consensus ranking

060 GitHub stars

Use Cases

01Automating the identification and resolution of complex logic bugs and performance bottlenecks

02Performing deep architectural and security audits on critical modules before production

03Executing pre-commit reviews to ensure high-consensus quality standards across the team

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dev-jelly/jelly-dotclaude jelly-multi-ai-code-review

For use in Claude.ai and ChatGPT

Download Skill