Website to Markdown

This MCP server fetches content from websites and converts it into Markdown with AI-powered cleanup, ready for AI processing, analytics, and rich output formats. It supports multi-site configurations and can be run locally or via an immediate NPX one-shot setup, making it easy to process API docs, blogs, and other web content into clean Markdown.

How to use

You use this MCP server by running it as a local tool within your MCP client workflow. Create a configuration that lists the websites you want to fetch, then start the MCP server and issue a command to process those sites. The server provides enhanced Markdown output with a content summary, reading time, word count, language detection, and a content quality score. When you query the server, you can instruct it to fetch a site, convert the content to Markdown, and retrieve the enriched metadata.

How to install

Prerequisites you need before installation:

• • Node.js 20.18.1+ (recommended: Node.js v22.15.0 LTS)
• • npm 10.0.0+ or yarn
• • Cursor Editor (optional for the MCP workflow)

Option A: NPX installation (recommended for quick start)

# Step 1: Create configuration file
# Create a file named my-websites.json listing the sites to fetch
{
  "websites": [
    {
      "name": "your_website",
      "url": "https://your-website.com",
      "description": "Your Project Website"
    },
    {
      "name": "api_docs",
      "url": "https://api.example.com/openapi.json",
      "description": "Your API Specification"
    }
  ]
}

# Step 2: Configure MCP Server
# Add to the MCP configuration with an environment variable pointing to your config file
{
  "mcpServers": {
    "website_md_mcp": {
      "command": "npx",
      "args": ["-y", "website-to-markdown-mcp"],
      "disabled": false,
      "env": {
        "WEBSITES_CONFIG_PATH": "./my-websites.json"
      }
    }
  }
}

# Step 3: Restart Cursor and test

Option B: Local installation (for development or customization)

# Step 1: Clone the project
# git clone https://github.com/your-username/website-to-markdown-mcp.git
# Step 2: Install dependencies
# npm install
# Step 3: Build project
# npm run build

# Step 4: Configure MCP Server
# Add to the MCP configuration with an environment variable pointing to your config file
{
  "mcpServers": {
    "website_md_local": {
      "command": "cmd",
      "args": ["/c", "node", "./website-to-markdown-mcp/dist/index.js"],
      "disabled": false,
      "env": {
        "WEBSITES_CONFIG_PATH": "./my-websites.json"
      }
    }
  }
}

Additional notes

You can place the configuration file anywhere accessible by the MCP server. Ensure the path you reference exists and is readable by the running process. The server will use the JSON configuration to fetch and convert content from the listed websites.

Security and best practices

• Use HTTPS URLs whenever possible to securely fetch content.

• Limit the amount of content processed in a single run to avoid long processing times.

• Review and sanitize the resulting Markdown to remove any extraneous content that may be unsuitable for your target environment.

Available tools

fetch_website

Fetches the HTML content of a website by URL.

list_configured_websites

Lists all configured websites from the local configuration.

fetch_petstore_openapi

Fetch Petstore OpenAPI 3.0 specification.

fetch_petstore_swagger

Fetch Petstore Swagger 2.0 specification.

fetch_github_api

Fetch GitHub API OpenAPI description.

fetch_tailwind_css

Fetch Tailwind CSS documentation OpenAPI/Markdown content.