Gemini Collaboration

An MCP server that allows Claude Code orchestrate Gemini for analysis and context-heavy tasks.

javascript

25

GitHub Stars

javascript

Language

4 months ago

First Indexed

3 weeks ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "dnnyngyen-gemini-cli-orchestrator": {
      "command": "node",
      "args": [
        "/absolute/path/to/gemini-cli-orchestrator/gemini-collaboration-guide.mjs"
      ]
    }
  }
}

You have a metaprompting-driven MCP server that guides AI agents through multi-step analysis of a codebase using Gemini’s capabilities. It exposes a lightweight, multi-tool workflow that helps you plan, craft prompts, iterate, and synthesize findings, enabling structured, stepwise reasoning without replacing your own thinking.

How to use

Use a compatible MCP client to connect and run the Gemini collaboration workflow. You will plan an analysis, craft tailored prompts for each step, iterate when needed, and synthesize the results into a final report. The server provides four dedicated tools that guide your analysis in a clear, repeatable sequence.

How to install

Prerequisites: Node.js 18+ and npm installed on your system.

Install the Gemini CLI globally if you do not have it already.

npm install -g @google/gemini-cli

Set up and test the MCP server integration by installing dependencies for the project and validating the entry point.

npm install
node gemini-collaboration-guide.mjs

Additional sections

MCP configuration uses a local stdio server that runs the entry point via Node. The server is configured to execute the script that coordinates Gemini-driven prompts.

If you use Claude Code as your MCP client, you can register the local server with a precise path to the orchestrator script and verify it via the MCP list command, then start using the MCP tools.

Important notes: authentication is handled automatically by the Gemini CLI, so you don’t need extra credentials in the MCP setup.

Troubleshooting tips cover common issues like filename mismatches, missing paths, and ensuring Claude Code restarts after configuration changes.

Tools available through the MCP server

gemini_plan_analysis goal: Breaks down complex analysis goals into a step-by-step plan. Be specific about what you want to understand and why.

gemini_craft_prompt step_description context: Helps you write better prompts for Gemini by suggesting effective commands and context for each analysis step.

gemini_iterate_analysis current_understanding iteration_goal: Guides iterative analysis using observe-think-act cycles for dynamic problem-solving.

gemini_synthesize_findings steps_summary synthesis_goal: Combines insights from multiple analysis steps into a comprehensive understanding and final recommendations.

Notes on configuration and troubleshooting

If you previously configured this MCP server or encounter issues like a failed status, clean the slate by removing old configurations and re-adding the server using the correct script path.

Verify the script file exists at the indicated absolute path, test the start command, and restart the MCP client after any configuration changes.

Available tools

gemini_plan_analysis

Breaks down complex analysis goals into a step-by-step plan. Be specific about what you want to understand and why.

gemini_craft_prompt

Helps you write better prompts for Gemini by suggesting effective commands and context for each analysis step.

gemini_iterate_analysis

Guide iterative analysis using observe-think-act cycles for dynamic problem-solving.

gemini_synthesize_findings

Combine insights from multiple analysis steps into comprehensive understanding and final recommendations.