GemForge (Gemini Tools)

Overview

GemForge-Gemini-Tools-MCP: Enterprise-grade Gemini integration for your favorite MCP agents. Supercharge Claude, Roo Code, and Windsurf with codebase analysis, live search, text/PDF/image processing, and more.

Quick Navigation

Features
Quick Start
Configuration
Tools
Heavy-Duty Reliability
Deployment
Examples
Community
Documentation

Why GemForge?

GemForge is the essential bridge between Google's Gemini AI and the MCP ecosystem: gemfog

Real-Time Web Access: Fetch breaking news, market trends, and current data with gemini_search
Advanced Reasoning: Process complex logic problems with step-by-step thinking via gemini_reason
Code Mastery: Analyze full repositories, generate solutions, and debug code with gemini_code
Multi-File Processing: Handle 60+ file formats including PDFs, images, and more with gemini_fileops
Intelligent Model Selection: Automatically routes to optimal Gemini model for each task
Enterprise-Ready: Robust error handling, rate limit management, and API fallback mechanisms

Quick Start

One-Line Install

npx @gemforge/mcp-server@latest init

Manual Setup

Create configuration file (claude_desktop_config.json):

{
  "mcpServers": {
    "GemForge": {
      "command": "node",
      "args": ["./dist/index.js"],
      "env": {
        "GEMINI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Install and run:

npm install gemforge-mcp
npm start

Watch 30-second setup demo →

Heavy-Duty Reliability

GemForge is built for production environments:

Support for 60+ File Types: Process everything from code to documents to images
Automatic Model Fallbacks: Continues functioning even during rate limits or service disruptions
Enterprise-Grade Error Logging: Detailed diagnostics for troubleshooting
API Resilience: Exponential backoff, retry logic, and seamless model switching
Full Repository Support: Analyze entire codebases with configurable inclusion/exclusion patterns
XML Content Processing: Specialized handling for structured data

Key Tools

Tool	Description	Key Capability
`gemini_search`	Web-connected information retrieval	Real-time data access
`gemini_reason`	Complex problem solving with step-by-step logic	Transparent reasoning process
`gemini_code`	Deep code understanding and generation	Full repository analysis
`gemini_fileops`	Multi-file processing across 60+ formats	Document comparison and transformation

Example: Real-Time Search

{
  "toolName": "gemini_search",
  "toolParams": {
    "query": "Latest advancements in quantum computing",
    "enable_thinking": true
  }
}

Example: Code Analysis

{
  "toolName": "gemini_code",
  "toolParams": {
    "question": "Identify improvements and new features",
    "directory_path": "path/to/project",
    "repomix_options": "--include \"**/*.js\" --no-gitignore"
  }
}

Example: Multi-File Comparison

{
  "toolName": "gemini_fileops",
  "toolParams": {
    "file_path": ["contract_v1.pdf", "contract_v2.pdf"],
    "operation": "analyze",
    "instruction": "Compare these contract versions and extract all significant changes."
  }
}

Configuration

GemForge offers flexible configuration options:

Environment Variables

GEMINI_API_KEY=your_api_key_here       # Required: Gemini API key
GEMINI_PAID_TIER=true                  # Optional: Set to true if using paid tier (better rate limits)
DEFAULT_MODEL_ID=gemini-2.5-pro        # Optional: Override default model selection
LOG_LEVEL=info                         # Optional: Set logging verbosity (debug, info, warn, error)

Claude Desktop Integration

{
  "mcpServers": {
    "GemForge": {
      "command": "node",
      "args": ["./dist/index.js"],
      "env": {
        "GEMINI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Advanced Model Selection

GemForge intelligently selects the best model for each task:

gemini_search: Uses gemini-2.5-flash for speed and search integration
gemini_reason: Uses gemini-2.5-pro for deep reasoning capabilities
gemini_code: Uses gemini-2.5-pro for complex code understanding
gemini_fileops: Selects between gemini-2.0-flash-lite or gemini-1.5-pro based on file size

Override with model_id parameter in any tool call or set DEFAULT_MODEL_ID environment variable.

Deployment

Smithery.ai

One-click deployment via Smithery.ai

Docker

docker run -e GEMINI_API_KEY=your_api_key ghcr.io/pv-bhat/gemforge:latest

Self-Hosted

Use our MCP.so Directory listing for integration instructions.

What Sets GemForge Apart?

Cross-Ecosystem Power: Bridge Google's AI with Claude and other MCP agents
Multi-File Analysis: Compare documents, images, or code versions
Smart Routing: Automatic model selection based on task requirements
Production-Ready: Built for enterprise environments

GemForge in Action

Community & Support

Join Us: MCP Discord | GemForge Discord
Contribute: GitHub Discussions
Feedback: Open an issue or share thoughts on Discord

Documentation

Visit our Documentation Site for:

Advanced usage tutorials
API reference
Troubleshooting tips

License

Licensed under the MIT License. See LICENSE for details.

Acknowledgments

Google Gemini API for providing the underlying AI capabilities
Model Context Protocol (MCP) for standardizing AI tool interfaces

GemForge Gemini