Cloudflare Edge Services

🚀 CF-MCP-Server

An image generation service implementation based on the Model Context Protocol (MCP), built with Cloudflare Workers.

🚀 Quick Start

Environment Requirements

• Node.js version 18 or higher
• Wrangler CLI
• NPM package management tool

Installation and Deployment

# Install dependencies
pnpm install

# Configure environment variables
cp .env.example .env

# Deploy the Worker
pnpm deploy

✨ Features

• 🖼️ Standardized service interface based on the MCP protocol
• 🚀 Ready-to-use image generation Worker module
• 🔧 Support for rapid expansion of custom tools and resources
• ⚡ Optimized deployment for edge computing

📦 Module Composition

mcp-image-generation

graph LR
    subgraph mcp-image-generation
    A[Cloudflare Worker] --> B[Flux-1-Schnell model]
    A --> C[Image generation API]
    C --> D((4-8 steps))
    end

• Supported step range: 4 - 8 steps
• Response format: Base64-encoded PNG image
• Deployment method: Cloudflare Workers

💻 Usage Examples

Image Generation API

Configuration Example:

{
  "mcpServers": {
    "generateImage": {
      "command": "/Users/zhangxudong/Gits/turinhub/cf-mcp-server/mcp-image-generation/node_modules/.bin/workers-mcp",
      "args": [
        "run",
        "generateImage",
        "https://mcp-image-generation.turinhub.com",
        "/Users/zhangxudong/Gits/turinhub/cf-mcp-server/mcp-image-generation"
      ],
      "env": {}
    }
  }
}

Parameter Description:

📚 Documentation

Adding New Tools

Refer to the online documentation Official example code repository

1. Create a new Worker using the MCP SDK:

npx create-cloudflare@latest my-new-worker
cd my-new-worker # I always forget this step
npm install workers-mcp
npx workers-mcp setup

⚠️ Important Note

Do not install wrangler in the project's local package.json. Use global management directly.

2. Update the main service configuration after implementing the tool logic.
3. Test MCP using @modelcontextprotocol/inspector:

npx @modelcontextprotocol/inspector

🤝 Contribution Guidelines

Contributions through Issues and PRs are welcome. Please follow the existing code style.

📄 License

MIT License © 2025 Turinhub