Puppeteer MCP Server

Browser automation and web scraping via MCP — let agents control Chrome

By Anthropic (official) ★ 4.3K stars official browser scraping

The Puppeteer MCP Server gives AI agents a full headless Chrome browser. Agents can navigate websites, fill forms, click buttons, take screenshots, and extract structured data from pages that JavaScript renders dynamically.

Installation & Configuration

Add this block to your claude_desktop_config.json (or your agent framework's MCP config file):

{
  "mcpServers": {
    "puppeteer": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-puppeteer"]
    }
  }
}

Config file location:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/claude/claude_desktop_config.json

Available Tools (6)

puppeteer_navigate
puppeteer_screenshot
puppeteer_click
puppeteer_type
puppeteer_select
puppeteer_evaluate

Use Cases

Web scraping — extract data from dynamic JavaScript websites
UI testing — have Claude write and run browser tests
Screenshots — capture pages for visual review

Source & Documentation

GitHub Repository: https://github.com/modelcontextprotocol/servers/tree/main/src/puppeteer

Protocol: Model Context Protocol (MCP) — modelcontextprotocol.io

← Back to MCP Server Directory