agents-libagents-lib
AgentsMCP ServersWorkflowsTutorialsComparisonsTemplatesNewsAbout
🔌

Fetch MCP

WebUtils•⭐ 900

Efficiently fetch and convert web pages to clean Markdown for LLM consumption.

GitHubWebsite →
Claude DesktopCursorWindsurf

Overview

Efficiently fetch and convert web pages to clean Markdown for LLM consumption.

Setup

Run with npx:

npx -y @modelcontextprotocol/server-fetch

Configuration

Added to claude_desktop_config.json

Documentation

Fetch MCP Server

Overview

The Fetch MCP Server provides a specialized tool for AI agents to retrieve the content of web pages in a format optimized for LLM consumption. Instead of returning raw HTML, which is often cluttered with scripts, styles, and boilerplate, the Fetch server converts web pages into clean, structured Markdown.

This allows agents to read web content more efficiently, reducing token usage and improving the quality of the analysis by removing irrelevant noise from the page.

Features

  • HTML to Markdown Conversion: Automatically converts complex HTML pages into clean Markdown, preserving the structure and key content.
  • Noise Reduction: Strips away navigation menus, footers, and other boilerplate content that doesn't contribute to the page's primary information.
  • Optimized for LLMs: Produces output that is easier for models to parse and summarize.
  • Fast Retrieval: Efficiently fetches page content and performs the conversion in real-time.

Installation

# Run via npx (recommended for MCP)
npx -y @modelcontextprotocol/server-fetch

Configuration

To use the Fetch MCP server, add it to your claude_desktop_config.json:

{
  "mcpServers": {
    "fetch": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-fetch"
      ]
    }
  }
}

Available Tools

ToolDescription
fetchTakes a URL and returns the cleaned Markdown content of the page.

Usage Examples

Researching a Topic

The agent can use the fetch tool to gather information from several official documentation pages, combining the clean Markdown outputs to synthesize a comprehensive answer.

Summarizing a News Article

The agent can fetch a long news article and immediately pass the cleaned Markdown to the LLM for summarization, avoiding the need to manually strip HTML tags.

Pros

  • ✅ Token Efficiency: significantly reduces the number of tokens used by removing HTML noise.
  • ✅ Higher Accuracy: LLMs perform better on structured Markdown than on raw, messy HTML.
  • ✅ Ease of Use: Single-tool interface for a critical agentic capability.

Cons

  • ❌ Static Content Only: Unlike Puppeteer, the Fetch server typically cannot interact with pages that require JavaScript execution to display content.
  • ❌ Potential Content Loss: In rare cases, the conversion process might strip out important but non-standard elements.

Resources