Read Website Fast

You can extract web content quickly and efficiently with the Read Website Fast MCP server. It fetches pages locally, strips noise, converts them to clean Markdown, and preserves links for easy use in AI workflows and knowledge graphs.

How to use

Use Read Website Fast to fetch a webpage and convert it to clean Markdown for downstream processing. You can run the MCP server through your preferred MCP client, add it to your client’s MCP configuration, and then request a page to be converted. The primary available tool is read_website, which takes a URL and returns the page content in Markdown (with optional metadata). You can also access cached results and cache statistics via the provided endpoints.

How to install

Prerequisites: ensure you have Node.js installed (LTS version recommended) and a supported MCP client that can consume MCP servers. You will primarily use npx to run the MCP from the command line.

claude mcp add read-website-fast -s user -- npx -y @just-every/mcp-read-website-fast

code --add-mcp '{"name":"read-website-fast","command":"npx","args":["-y","@just-every/mcp-read-website-fast"]}'

cursor://anysphere.cursor-deeplink/mcp/install?name=read-website-fast&config=eyJyZWFkLXdlYnNpdGUtZmFzdCI6eyJjb21tYW5kIjoibnB4IiwiYXJncyI6WyIteSIsIkBqdXN0LWV2ZXJ5L21jcC1yZWFkLXdlYnNpdGUtZmFzdCJdfX0=

{
  "mcpServers": {
    "read-website-fast": {
      "command": "npx",
      "args": ["-y", "@just-every/mcp-read-website-fast"]
    }
  }
}

Notes on usage patterns

After adding Read Website Fast to your MCP client, you can request a page to be fetched and converted by invoking the read_website tool with a URL. You can tune how many pages to crawl, control concurrency, and decide output format (Markdown, JSON metadata, or both) according to your client’s capabilities. Caches reduce token usage on repeated fetches, and robots.txt support helps you crawl pages politely.

Additional setup and endpoints

You can check cache statistics or clear the cache using the dedicated endpoints exposed by the MCP server setup. These endpoints help you monitor performance and reset local caches when needed.

Examples of common actions

• Fetch a single page and get Markdown output using the default settings. - Fetch with multiple pages by setting the pages limit. - Retrieve JSON metadata or both Markdown and metadata for richer downstream processing.

Available tools

read_website

Fetches a webpage and converts it to clean Markdown with preserved links. Optional pages parameter controls how many pages to crawl.

read_website_status

Provides cache statistics for the Read Website Fast MCP server.

read_website_clear_cache

Clears the local cache directory used for storing fetched pages and MT metadata.