Puppeteer MCP Server

Overview

Puppeteer MCP Server provides browser automation capabilities through the standardized Model Context Protocol framework. This implementation harnesses Puppeteer's powerful web automation library to enable AI systems to navigate websites, interact with page elements, capture screenshots, and execute JavaScript. By providing these capabilities, the server bridges the gap between language models and real-world web interaction, enabling sophisticated web-based workflows and operations.

Features

• Browser Environment Access: Interact with websites in a real browser environment
• Element Interaction: Click, hover, fill forms, and select dropdown options
• Screenshot Capabilities: Capture full-page or element-specific screenshots
• JavaScript Execution: Run custom scripts within the browser context
• Console Monitoring: Access and monitor browser console logs
• Full Navigation Control: Browse the web with complete page loading support
• Multiple Runtime Options: Choose between headless (Docker) or visible browser (NPX) environments

Use cases

• Building web scraping solutions that require JavaScript execution
• Creating automated testing frameworks for web applications
• Generating website screenshots for documentation or monitoring
• Developing interactive demos with web-based user interfaces
• Automating form submission and data entry processes
• Enabling AI assistants to interact with web applications
• Monitoring website changes and performance metrics
• Performing visual regression testing across web interfaces
• Creating visual content from web sources

Tools

• puppeteer_navigate: Navigate browser to any URL
• puppeteer_screenshot: Capture screenshots with customizable parameters
• puppeteer_click: Click on page elements using CSS selectors
• puppeteer_hover: Hover over page elements
• puppeteer_fill: Fill out form fields with specified values
• puppeteer_select: Make dropdown selections
• puppeteer_evaluate: Execute JavaScript in the browser context

Integration complexity

3/5 - Moderate complexity requiring basic configuration choices and understanding of browser automation concepts.

Setup Guide

Docker Configuration (Headless Mode)

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "puppeteer": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "--init",
        "-e",
        "DOCKER_CONTAINER=true",
        "mcp/puppeteer"
      ]
    }
  }
}

NPX Configuration (Visible Browser)

Add to your claude_desktop_config.json:

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

Building From Source

To build a Docker image:

docker build -t mcp/puppeteer -f src/puppeteer/Dockerfile .

License

MIT License

Website

Product: https://pptr.dev/ Github: https://github.com/modelcontextprotocol/servers/tree/main/src/puppeteer