When should I use this skill instead of standard Claude Code?

This skill is best for complex features, security-sensitive code, or large refactors where an extra layer of automated validation is needed to ensure high quality.

What is the Gemini-Claude Engineering Loop?

It is a specialized workflow for Claude Code that uses Gemini as a secondary reviewer to validate Claude's implementation plans and final code output.

Does it store a history of the AI collaboration?

Yes, all plans, validation results, and review histories are stored locally in the .gemini-loop directory for your reference.

Can I choose which Gemini model to use for reviews?

Yes, the skill allows you to select between different models, such as gemini-3-flash-preview for speed or gemini-3-pro-preview for more complex architectural reviews.

Does this skill require a Gemini API key?

Yes, this skill depends on the gemini-cli plugin, which requires a configured Google Gemini API key to function.

Gemini-Claude Engineering Loop

Name: Gemini-Claude Engineering Loop
Author: creator-hian

bycreator-hian

•

Productivity & Workflow

Orchestrates a dual-AI engineering workflow where Claude manages implementation and Gemini provides rigorous validation and code review.

The Gemini-Claude Engineering Loop skill introduces a sophisticated development cycle that leverages the complementary strengths of two leading AI models. In this workflow, Claude acts as the architect and implementer, while Gemini serves as the quality assurance and validation engine. By alternating between Claude's planning/coding and Gemini's review/critique phases, this skill significantly reduces architectural errors, enhances security, and ensures high-performance code for complex features or large-scale refactoring projects.

Key Features

01Automated dual-AI feedback loops for planning and implementation

02Persistent context tracking via a structured .gemini-loop directory

032 GitHub stars

04Independent plan validation by Gemini to catch architectural flaws early

05Configurable role modes for review-only or suggestion-based feedback

06Automated post-implementation code reviews for quality assurance

Use Cases

01Mission-critical engineering tasks requiring a secondary AI audit

02Developing complex features that require multi-stage validation

03High-stakes refactoring where security and performance are critical

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add creator-hian/claude-code-plugins gemini-claude-loop

For use in Claude.ai and ChatGPT

Download Skill