home / mcp / mcp website downloader mcp server

MCP Website Downloader MCP Server

MCP tool for Claude to download entire Windows working website url and assets and save to a library for AI use.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "angrysky56-mcp-windows-website-downloader": {
      "command": "python",
      "args": [
        "-m",
        "mcp_windows_website_downloader.server",
        "--library",
        "docs_library"
      ],
      "env": {
        "PYTHONUNBUFFERED": "1"
      }
    }
  }
}

You have a lightweight MCP server that downloads documentation websites and prepares them for RAG indexing. It keeps the download process simple, organizes assets, and exposes a straightforward way to run the downloader locally as an MCP tool.

How to use

To use the MCP server, run it as a local process and interact with it through your MCP client. Start by launching the server, then issue a download command for a documentation site. The server will fetch the site, preserve its structure, collect assets, and generate a rag_index.json describing the downloaded domain and pages.

How to install

Prerequisites: Python must be installed on your system. You will also need a Python virtual environment tool (venv) and permission to install Python packages.

Step-by-step commands you should run in order.

# Create and activate a virtual environment
python -m venv venv
# On Windows
venv\Scripts\activate
# On Unix or macOS
source venv/bin/activate
# Install the MCP package in editable mode
pip install -e .

Additional sections

Configuration and running details are shown below. You can run the server locally to fetch a documentation site and generate an index suitable for RAG workflows.

Starting the server in development mode uses Python to run the MCP server module with a library path for downloaded sites.

python -m mcp_windows_website_downloader.server --library docs_library