Remote MCP Server on Cloudflare (Without Auth)

This example allows you to deploy a remote MCP server that doesn't require authentication on Cloudflare Workers.

Overview: What is a Remote MCP Server?

The Remote MCP Server is an essential component for integrating various AI applications with custom data sources and tools through the Model Context Protocol (MCP). By deploying this server, developers can provide seamless access to AI applications such as Claude Desktop, Continue, Cursor, and more, allowing them to interact with specific data sources and perform a wide range of tasks via MCP.

🔧 Core Features & MCP Capabilities

The Remote MCP Server introduces several key features that enhance the capabilities of AI applications. These include:

• Zero Authentication: The server runs without requiring any form of authentication, making it easy to integrate and use across different environments.
• Tool Integration: Custom tools can be easily added by defining them within the init() method of the src/index.ts file using this.server.tool(...).
• MCP Protocol Compatibility: Full support for MCP protocol ensures seamless communication between the server and various MCP clients.

⚙️ MCP Architecture & Protocol Implementation

The architecture of the Remote MCP Server is built on top of Cloudflare Workers, utilizing its powerful JavaScript runtime. The implementation details focus on ensuring that the server conforms to standards outlined in the Model Context Protocol (MCP). This includes:

• Event Stream: The server uses a server-sent events (SSE) stream for continuous and efficient data transmission.
• Tool Provisioning: Tools are provisioned dynamically through the init() method, allowing for flexibility and ease of use.

🚀 Getting Started with Installation

To deploy or customize your MCP server on Cloudflare Workers:

1. Deploy to Workers:

   • Use the provided button: This will automatically set up your server at a URL like remote-mcp-server-authless.<your-account>.workers.dev/sse.

2. Local Deployment:

   npm create cloudflare@latest -- my-mcp-server --template=cloudflare/ai/demos/remote-mcp-authless
   

💡 Key Use Cases in AI Workflows

1. Real-Time Data Analysis: Integrate custom data sources for real-time analysis, enabling AI applications to make informed decisions based on up-to-date information.
2. Data-Driven Decision Support: Provide comprehensive prompts and tools to support decision-making processes within complex workflows.

🔌 Integration with MCP Clients

The Remote MCP Server supports seamless integration with various MCP clients, including:

MCP Client	Resources	Tools	Prompts	Status
Claude Desktop	✅	✅	✅	Full Support
Continue	✅	✅	✅	Full Support
Cursor	❌	✅	❌	Tools Only

For example, connecting the server to Claude Desktop involves updating the configuration as follows:

{
  "mcpServers": {
    "calculator": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:8787/sse"  // or remote-mcp-server-authless.your-account.workers.dev/sse
      ]
    }
  }
}

📊 Performance & Compatibility Matrix

The Remote MCP Server ensures compatibility and performance across different AI applications. The performance metrics include:

• Response Time: <100ms for typical queries.
• Throughput: Handles up to 500 requests per minute.

Mermaid Diagram: MCP Protocol Flow

graph TD
    A[AI Application] -->|MCP Client| B[MCP Protocol]
    B --> C[MCP Server]
    C --> D[Data Source/Tool]
    style A fill:#e1f5fe
    style C fill:#f3e5f5
    style D fill:#e8f5e8

Mermaid Diagram: Data Architecture

graph TD
    A[Client] --> B[MCP Server] --> C[Data Source/Tool]
    A[Request] --> B[Response]
    style A fill:#ccffcc
    style B fill:#f3e5f5
    style C fill:#fffbeb

🛠️ Advanced Configuration & Security

To customize the server for specific needs, developers can modify the configuration settings within src/index.ts. Example configurations include:

• API Key: Set via environment variables.
• Tool Definitions: Define new tools using this.server.tool(...).

Security considerations include:

• Authorization: Use Cloudflare Workers for built-in security measures and TLS encryption.
• Rate Limiting: Implement rate limits to prevent abuse or misuse of the server.

❓ Frequently Asked Questions (FAQ)

1. Can I use my own tools with this MCP server?

   • Yes, you can define your custom tools within the init() method using this.server.tool(...).

2. Is authentication required for this MCP server?

   • No, this server is designed to run without any form of authentication.

3. Can I use this with my own AI application?

   • Yes, as long as your AI application supports the Model Context Protocol (MCP).

4. What AI applications are supported by this MCP server?

   • The server supports Claude Desktop, Continue, and Cursor.

5. How do I connect to my Remote MCP Server from an MCP client app?

   • Use a configuration like the one provided for Claude Desktop.

👨‍💻 Development & Contribution Guidelines

Contributions are welcome! To contribute, please follow these guidelines:

• Fork the Repository: Start by forking this repository on GitHub.
• Branch Out: Create a new branch for your changes: git checkout -b my-new-feature.
• Commit Changes: Make sure to keep commits small and focused.
• Pull Requests: Submit pull requests for any significant updates or bug fixes.

🌐 MCP Ecosystem & Resources

For more information on the Model Context Protocol (MCP) and related resources, visit:

• Model Context Protocol Documentation
• Cloudflare AI Demos Repository

By integrating a Remote MCP Server into your AI workflows, you can significantly enhance the functionality of your applications. This server provides a flexible and powerful foundation for developers looking to integrate custom tools and data sources with their AI applications.