MCP Node Fetch

MCP Node Fetch is an MCP server that enables language model assistants to retrieve web content and query remote endpoints using the Node.js undici HTTP library. It bridges the gap between desktop AI clients and the public internet, allowing assistants to perform network requests on demand. Developers, technical researchers, and automated workflow designers use this tool to bring external web resources directly into conversational contexts. By connecting to the high-performance undici client, the server facilitates rapid retrieval of raw HTML documents, external REST API payloads, and plain text files without requiring browser automation overhead. Users can prompt their assistant to inspect web pages, pull current remote data, or review online technical documentation in real time. Because it operates through standard Model Context Protocol interfaces, client applications can trigger network calls systematically during planning and execution phases. The server simplifies development pipelines by providing a focused, lightweight HTTP retrieval capability that avoids the resource constraints of running full headless browser instances for basic fetching tasks.

Category: Browser & Web Automation

Tags: fetch, http, nodejs, undici, web scraping

Visit MCP Node Fetch

How to install and configure MCP Node Fetch

1. Consult the project hosting page at https://mcpservers.org/servers/mcollina/mcp-node-fetch to identify the recommended installation package and repository instructions. 2. Ensure that a supported Node.js runtime environment is installed on your system. 3. Open your MCP client configuration file, such as the Claude Desktop configuration JSON. 4. Register the server under your client's mcpServers section using the designated node launch command. 5. Save the configuration file and completely restart your MCP client to verify tool availability.

What you can do with MCP Node Fetch

• Scrape website content: Download and parse remote HTML pages to extract structured article text, documentation, or metadata directly into LLM prompts. * Query REST APIs: Send HTTP GET requests to third-party endpoints to fetch JSON data payloads and analyze live service metrics. * Inspect raw response headers: Read status codes, content-type headers, and network response metadata for debugging web endpoints during development. * Retrieve remote documentation: Fetch developer guides, API specs, and technical documentation directly from the web for contextual coding assistance.

Key facts

• Browser & Web Automation, Developer Tools & Code Intelligence
• fetch, http, nodejs, undici, web scraping

Part of MCP Servers

What can MCP Node Fetch do?

MCP Node Fetch enables Model Context Protocol clients to issue HTTP GET requests across the internet using Node.js and the undici library. It allows AI models to download web page HTML, query RESTful APIs, inspect HTTP response headers, and incorporate remote online data into active chat conversations without launching heavy browser automation instances.

Which MCP clients work with MCP Node Fetch?

The server functions with any client that implements the Model Context Protocol standard over standard input and output streams. Compatible applications include Claude Desktop, Cursor, and custom agentic frameworks configured to run Node.js-based MCP processes locally on macOS, Linux, or Windows environments.

How do I install MCP Node Fetch?

Installation and launch instructions depend on the repository distribution. Users typically register the server in their MCP client configuration by referencing Node.js and the package name. Review the project documentation at the external server page to verify exact launch flags and environment prerequisites before adding the server to your settings.

Why does this server use the undici library?

Undici is an official, highly optimized HTTP/1.1 client written from scratch for Node.js. It offers significantly better performance, lower resource consumption, and improved reliability compared to legacy Node.js HTTP modules, making it a reliable choice for lightweight web retrieval within tool-calling environments.