Fetch MCP Server

What is Fetch MCP Server?

Fetch MCP Server is a Model Context Protocol (MCP) implementation with web content fetching capabilities, using Server-Sent Events (SSE) for real-time communication. It is designed for AI applications that need a standardized way to manage and communicate with models and resources.

How to use Fetch MCP Server?

Clone the repository, install dependencies with Yarn (yarn install), then start in development mode with yarn dev or production mode with yarn start. The server runs on port 3001 by default (overridable via the PORT environment variable). Two endpoints are provided: /sse for persistent SSE connections and /messages for sending messages to the MCP server.

Key features of Fetch MCP Server

• Model Context Protocol (MCP) implementation
• Server-Sent Events (SSE) for real-time communication
• Web content fetching with Playwright
• Google search capability with SERP results
• HTML to Markdown conversion
• Resource templating and management

Use cases of Fetch MCP Server

• Real-time communication between AI applications and models using SSE
• Fetching and converting web page content to Markdown for AI processing
• Performing web searches and retrieving structured SERP data
• Standardized resource management within an MCP ecosystem
• Deploying a scalable MCP server with Kubernetes

FAQ from Fetch MCP Server

What is Fetch MCP Server?

Fetch MCP Server is a Model Context Protocol server that provides web content fetching and search capabilities, using SSE for real-time communication. It is intended for AI applications that need a standardized way to interact with models and resources.

What are the prerequisites for running Fetch MCP Server?

Node.js version 14 or higher and Yarn package manager are required. A Kubernetes cluster is needed only if you plan to deploy using the provided Kubernetes configurations.

How do I start the Fetch MCP Server?

Run yarn dev for development mode (with MCP inspector) or yarn start for production mode. You can change the port by setting the PORT environment variable (e.g., PORT=5000 yarn start).

What tools are available in Fetch MCP Server?

The server exposes two tools: fetchUrls to fetch multiple URLs and convert their content to markdown, and search to retrieve search results from web pages with SERP results.

Does Fetch MCP Server require a Kubernetes cluster to run?

No, Kubernetes is only required for deploying the server in a cluster. The server can be run directly with Node.js and Yarn.