
    ==================
      
       /// MCP ///
      /// FET ///
        
    ==================
        
    [server:online]
    [protocol:ready]

fetcher-mcp

by jae-jae

MCP server that fetches web-page content via a Playwright headless browser, with Readability extraction and HTTP/SSE transports.

790

Open Source

Retrieve web page content from a specified URL using a Playwright headless browser with optional main-content extraction and HTML/Markdown output.

Fetch web page content from multiple URLs in parallel, combining results while supporting the same parameters as fetch_url.

Installation

1. Prerequisites

• Node.js ≥ 18

• npm or pnpm/yarn

• (First-time only) Install Playwright browsers: `npx playwright install --with-deps`

2. Clone & install

   git clone https://github.com/jae-jae/fetcher-mcp.git
   cd fetcher-mcp
   npm install        # or pnpm i / yarn

3. Configure

• Copy the sample environment file (if present): `cp .env.example .env`

• Set any needed variables, e.g.

PLAYWRIGHT_CHROMIUM_EXECUTABLE=/usr/bin/chromium

PORT=8080

4. Build (TypeScript → JS)

npm run build # usually: tsc -p .

5. Start the MCP server

npm start # or: node dist/index.js

The service will listen on `http://localhost:8080` (or PORT).

6. (Optional) Docker

   docker build -t fetcher-mcp .
   docker run -p 8080:8080 fetcher-mcp

License: MIT License

Updated 7/30/2025