Puppeteer MCP Quick Start

Puppeteer MCP Quick Start

Version: 1.0.13
Reading Time: 2 minutes

Welcome to Puppeteer MCP! 🚀

Get started with browser automation in minutes. Choose your path:

🎯 Quick Navigation

📦 Installation

All installation methods - npm, source, Docker
2 minute setup

🚀 First Steps

Basic usage examples and your first automation
5 minute tutorial

🤖 Claude Desktop

Set up AI-powered browser automation
3 minute setup

⚙️ Configuration

Essential settings and environment setup
5 minute guide

What is Puppeteer MCP?

Puppeteer MCP is a beta browser automation platform that provides:

• Multi-Protocol Support: REST, gRPC, WebSocket, and Model Context Protocol (MCP)
• AI Integration: Works seamlessly with Claude Desktop and other AI assistants
• Enterprise Security: JWT + API key authentication, NIST compliance
• Comprehensive Automation: Full Puppeteer API with session management
• Zero Setup: Works out of the box with sensible defaults

Choose Your Starting Point

💡 I want to…

Use with Claude Desktop

1. Install globally: npm install -g puppeteer-mcp
2. Configure Claude Desktop (see guide)
3. Ask Claude to browse websites for you!

Integrate into my Node.js project

1. Install in project: npm install puppeteer-mcp
2. Import and configure (see examples)
3. Start automating browsers programmatically

Run as a standalone service

1. Clone repository or use Docker (see installation)
2. Configure environment (see configuration)
3. Access via REST API, gRPC, or WebSocket

Try it quickly without installation

npx puppeteer-mcp

This runs the latest version without installing anything!

Quick Example

Here’s browser automation in action:

// Using the REST API
const response = await fetch('http://localhost:8443/api/sessions', {
  method: 'POST',
  headers: {
    Authorization: 'Bearer your-token',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    baseUrl: 'https://example.com',
  }),
});

const { sessionId } = await response.json();

// Take a screenshot
await fetch(`http://localhost:8443/api/sessions/${sessionId}/screenshot`, {
  method: 'POST',
  headers: {
    Authorization: 'Bearer your-token',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    fullPage: true,
  }),
});

System Requirements

Minimum

• Node.js 18.0.0+
• 2GB RAM
• 1GB disk space

Recommended

• Node.js 20.0.0+
• 4GB RAM
• 2GB disk space
• Chrome/Chromium installed

Common Use Cases

• 🕷️ Web Scraping: Extract data from dynamic websites
• 🧪 Automated Testing: End-to-end browser testing
• 📄 PDF Generation: Convert web pages to PDFs
• 📸 Screenshots: Capture website screenshots
• 🤖 AI Automation: Let AI agents browse the web
• 📊 Monitoring: Track website changes and performance

📣 Beta Feedback Welcome!

As this is a beta release, we’re actively seeking your feedback:

• 🐛 Report Issues: Help us identify bugs
• 💡 Share Ideas: What features would make this production-ready for you?
• 📊 Performance: Let us know about any performance concerns
• 📝 Documentation: Tell us what needs better explanation

Your feedback is crucial for making this project truly production-ready!

Need Help?

• 📚 Full Documentation: Comprehensive guides and API reference
• 💬 GitHub Issues: Report bugs or request features
• 🔧 Troubleshooting: Common issues and solutions

Ready to Start?

📦 Install Now | 🚀 View Examples | 🤖 Setup Claude

---

Pro Tip: Using npx puppeteer-mcp is the fastest way to try it out - no installation needed! 🎉