Mult-Fetch

You can run and use a Mult Fetch MCP Server to enable AI assistants to fetch and extract web content through a standardized, transport-agnostic interface. It supports intelligent content extraction, chunked content retrieval, and optional browser-based interactions, making it easier to process large pages while preserving context.

How to use

You will run the MCP server as a local process and connect an MCP client to it using Standard Input/Output (Stdio). The server exposes tools that fetch web content in multiple formats, optionally extract meaningful content, and return metadata. In your interactions, you can request HTML, JSON, plain text, Markdown, or plaintext representations of a page, and you can enable content splitting to work with long documents in chunks. Use language settings to control English or Chinese responses, and enable debug output when you need deeper visibility into operations.

How to install

Prerequisites: you need Node.js installed on your system. Optional but recommended is a modern package manager such as pnpm or npm to install dependencies and run scripts.

Step 1: Install or run the MCP server via your preferred method.

Step 2: Start the MCP server using the provided stdio command. The recommended configuration uses npx to run the MCP server in-place.

Step 3: Connect your MCP client to the running server as described in the usage guidance.

Configuration and runtime notes

To run the server via a standard MCP client setup using npx, configure the MCP entry as follows.

{
  "mcpServers": {
    "mult-fetch-mcp-server": {
      "command": "npx",
      "args": ["@lmcc-dev/mult-fetch-mcp-server"],
      "env": {
        "MCP_LANG": "en"
      }
    }
  }
}

Security and troubleshooting notes

Enable debug mode only when you need detailed logs. Debug output is sent to STDERR and can help diagnose transport, fetch, or browser interactions. If you encounter proxy or network issues, try explicitly setting a proxy in your request, or enable browser mode for sites with advanced protections.

Available tools

fetch_html

Fetches HTML content from a URL and returns the content in a selectable format such as HTML, JSON, text, Markdown, or plaintext.

fetch_json

Fetches JSON data from a URL and returns the parsed JSON for further processing.

fetch_txt

Fetches plain text content from a URL, stripping HTML structure when possible.

fetch_markdown

Fetches content and renders it as Markdown for easy display and formatting.

fetch_plaintext

Fetches content and returns plain text after stripping HTML tags and formatting.

listResources

List available resources and provide access to project-related files and documentation.

readResource

Read a specific resource using a resource URI and return its contents.

listPrompts

List available prompt templates that guide content extraction or website interaction.

getPrompt

Retrieve a specific prompt template by name and show its arguments.