WebforAI Text Extractor

You can use the WebforAI Text Extractor MCP Server to pull plain text from any web page and receive clean, Markdown-formatted content ready for AI models. This server wraps WebforAI capabilities in an MCP-compatible interface, making it easy to feed web content into tools and assistants that support MCP.

How to use

Connect your MCP client to the deployed server to extract text from web pages. You can access the server remotely or run it locally for development. Use the server to request a page URL and receive Markdown-formatted output where links are plain text, tables are preserved as text, and images are hidden.

How to install

Prerequisites: you need Node.js and a package manager on your system. A local development workflow uses pnpm in the examples.

Step 1: Clone the MCP server repository and enter the project directory.

Step 2: Install dependencies.

Step 3: Start the development server.

Step 4: Access the local server at the specified local URL.

# Clone the project
git clone https://github.com/yutakobayashidev/webforai-mcp-server.git
cd webforai-mcp-server

# Install dependencies
pnpm install

# Start the development server
pnpm dev

# Access locally
# http://localhost:8787

Connecting to MCP Clients

You can connect using either a direct HTTP MCP endpoint or a local command that runs an MCP client. The following examples show a remote HTTP URL as well as a local execution configuration.

If you deploy to Cloudflare Workers, you will get a URL like this to use as the MCP endpoint: webforai-mcp-server.<your-account>.workers.dev/sse.

For a local development setup, you can use a configuration similar to this in an MCP client. The example assumes you run the MCP server locally on port 8787.

{
  "mcpServers": {
    "webforaiExtractor": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:8787/sse"  
      ]
    }
  }
}

Security and notes

Keep your MCP endpoints secure. Use access controls and avoid exposing private endpoints publicly. When testing, prefer local URLs and private Cloudflare deployments before moving to production.

Examples and troubleshooting

If you encounter errors while extracting text, verify that the target URL is reachable and that the server is running. Check the local development console for error messages and ensure dependencies are properly installed.

Available tools

extractWebPageText

Extracts plain text from a given web page URL and returns Markdown-formatted content with links converted to plain text, tables to plain text, and images hidden.