Fetch MCP Server

Web content retrieval as HTML, JSON, plain text, or Markdown for AI assistants.

Data last scanned 2 days ago · Reviewed 2 days ago

154 tools★ 3TypeScript

Overview

This server allows AI assistants to fetch and transform web content directly from URLs. It utilizes JSDOM for parsing HTML and TurndownService to convert webpages into Markdown, facilitating easier information consumption for models. Users can supply custom headers during requests to handle basic authentication or site-specific requirements. The server operates as an MCP toolset, exposing four primary functions: fetch_html, fetch_json, fetch_txt, and fetch_markdown. It does not maintain persistent resources, functioning instead as an on-demand bridge for web data extraction.

Our verdict

With a high-risk security score of 15/100, this server is currently intended for experimental use rather than mission-critical workflows. The primary security concerns stem from its lack of maintenance, evidenced by the fact that the last commit occurred 493 days ago, and the repository is managed by an independent community developer rather than an official organization. Prospective users should be aware that the code has not undergone editorial review and lacks clear legal terms, as no license was detected.

Supports conversion of web content directly into Markdown
Provides custom header support for request configuration
Maintains a minimal supply-chain with only four direct dependencies

High security risk due to long-term inactivity
No detected license creates legal ambiguity
Community-maintained without official vendor support

Setup

Claude Desktop

{
  "mcpServers": {
    "fetch-mcp": {
      "args": [
        "-y",
        "fetch"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
fetch_html	Fetch a website and return the content as HTML	low
fetch_json	Fetch a JSON file from a URL	low
fetch_txt	Fetch a website and return the content as plain text (no HTML)	low
fetch_markdown	Fetch a website and return the content as Markdown	low

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›Can I include custom headers in my web requests?

Yes, every tool provided by this server accepts an optional headers object to include custom headers in your requests.

›Does this server store data or persistent resources?

No, the server does not provide any persistent resources and is designed exclusively to fetch and transform web content on demand.

›What tools are available for processing webpage content?

You can use fetch_html for raw HTML, fetch_json for JSON files, fetch_txt for plain text without tags or scripts, and fetch_markdown to convert webpages into Markdown format.

›What libraries does the server use for processing content?

The server utilizes JSDOM for HTML parsing and text extraction, and it uses TurndownService for converting HTML to Markdown.