Remote Authless MCP Server on Cloudflare

This example allows you to deploy a remote MCP server that doesn't require authentication on Cloudflare Workers, providing a universal adapter for AI applications.

Overview: What is a Remote Authless MCP Server?

An MPC (Model Context Protocol) server enables AI applications like Claude Desktop, Continue, Cursor, and others to connect to specific data sources and tools via a standardized protocol. This remote server variant simplifies deployment without the need for authentication, making it accessible from various devices.

🔧 Core Features & MCP Capabilities

Real-Time Communication

The server uses Server-Sent Events (SSE) over HTTP/2 for real-time communication with AI applications, ensuring efficient data streaming and low latency.

Tool Integration

You can add your own tools to the MCP server by defining them inside the init() method of src/index.ts using this.server.tool(...). This allows you to extend the functionality of your remote server according to specific needs.

⚙️ MCP Architecture & Protocol Implementation

MCP Server Deployment

To deploy this server, you can utilize a pre-built template via NPM or the provided Cloudflare Workers button:

• Using NPM:

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

• Alternatively: Click here to automatically deploy the server.

Remote Client Connection

Once deployed, you can connect your AI applications to this MCP server using the following steps:

1. Connect to Cloudflare AI Playground:

   • Go to https://playground.ai.cloudflare.com/
   • Enter your deployed MCP server URL (remote-mcp-server-authless.<your-account>.workers.dev/sse)
   • Use your MCP tools directly from the playground!

2. Connect Claude Desktop:

   • Follow Anthropic's Quickstart guide.
   • In settings, navigate to Developer > Edit Config.
   • Update with this configuration:

     {
       "mcpServers": {
         "calculator": {
           "command": "npx",
           "args": [
             "mcp-remote",
             "http://localhost:8787/sse"
           ]
         }
       }
     }
     

   • Restart Claude and see your tools become available.

🚀 Getting Started with Installation

Prerequisites

Ensure you have the necessary development environment set up:

• Node.js and npm installed.
• Basic understanding of TypeScript and WebAssembly.

Steps to Deploy MCP Server

1. Clone Repository:

   git clone https://github.com/cloudflare/ai.git
   cd ai/demos/remote-mcp-authless
   

2. Initialize Project: Run the following command:

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

3. Deploy to Cloudflare Workers: Follow the link provided or use npm run deploy within your project directory.

💡 Key Use Cases in AI Workflows

Example 1: Financial Analysis Tool Integration

Imagine an AI application designed for financial analysts. By integrating a remote MCP server deployed on Cloudflare, you can provide real-time stock price updates and predictive analysis tools directly through the server.

• Implementation Steps:
  • Define data sources (stock prices) in the init() method of src/index.ts.
  • Use the mcp-remote command to connect the AI application to these tools.

Example 2: Customer Feedback Analysis

A customer service chatbot could use a remote MCP server to analyze and categorize customer feedback from social media. Tools like sentiment analysis could be added directly into the server.

• Implementation Steps:
  • Configure the chatbot to send text data queries through SSE to the MCU server.
  • Use pre-defined tool functions in init() for sending the analyzed results back.

🔌 Integration with MCP Clients

Compatibility Matrix

The following table shows the compatibility of various MCP clients with this remote server:

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

Configuration Sample

{
  "mcpServers": {
    "calculator": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:8787/sse"
      ],
      "env": {
        "API_KEY": "your-api-key"
      }
    }
  }
}

📊 Performance & Compatibility Matrix

Protocol Flow Diagram

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

Real-Time Communication

Real-time updates between the AI application and tools via SSE ensure efficient data streaming.

🛠️ Advanced Configuration & Security

Custom Tool Integration

You can extend the functionality of your remote server by defining new tools within the init() method. Each tool is added using this.server.tool(...).

Security Considerations

• Ensure that your MCP server implements appropriate security measures (e.g., HTTPS, environment variable protection).
• Use API keys and other security mechanisms to protect against unauthorized access.

❓ Frequently Asked Questions (FAQ)

1. How do I integrate tools into the remote MCP server?

   • Define custom utilities in src/index.ts using this.server.tool(). Each utility can perform specific tasks like data analysis or integration with external APIs.

2. What is the compatibility matrix for this remote MCP server?

   • Refer to the table above under "Integration with MCP Clients" to understand which clients are fully supported.

3. How do I connect the Cloudflare AI playground to my deployed server?

   • Enter your deployed URL (remote-mcp-server-authless.<your-account>.workers.dev/sse) in the AI Playground interface and use the available tools directly.

4. Can this remote MCP server be used with other AI frameworks besides Claude Desktop?

   • Yes, it supports multiple AI applications through the mcp-remote command-line utility or API keys.

5. What are some security best practices for deploying an MCP server on Cloudflare Workers?

   • Use HTTPS and strong encryption methods to secure data in transit.
   • Protect environment variables and sensitive information using secrets management tools provided by Cloudflare.

👨‍💻 Development & Contribution Guidelines

Setting Up the Local Environment

1. Install Dependencies:

   npm install
   

2. Run Locally:

   npm run start
   

3. Contribute to Open Source:

   • Fork and clone the repository.
   • Make changes as needed.
   • Submit a pull request for review.

🌐 MCP Ecosystem & Resources

Additional Resources

• Model Context Protocol (MCP) Documentation: ModelContextProtocol.io.
• Cloudflare Workers Documentation: Workers.cloudflare.com/docs.

Conclusion

Deploying a remote authless MCP server on Cloudflare provides a scalable and secure solution for integrating various AI applications with external tools and data sources. With flexibility in customization and wide compatibility, this server caters to developers looking to enhance their AI workflows through standardized protocols.

By following the detailed steps outlined here, you can set up your own MCP server and seamlessly integrate it into existing or new AI applications.