Puppeteer MCP Server

Official

Let your AI control a real browser — screenshot, click, and scrape

puppeteerbrowserscrapingautomationofficialchrome

The official Puppeteer MCP server gives your AI assistant full control of a Chrome browser via Puppeteer. Navigate to URLs, take screenshots, click elements, fill forms, and extract structured data from any web page. Runs locally on your machine — no headless browser service required. Ideal for web scraping, UI testing, and automating web-based workflows that would otherwise require manual interaction.

Install Config

Config file: ~/Library/Application Support/Claude/claude_desktop_config.json· Windows: %APPDATA%\Claude\claude_desktop_config.json

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

Available Tools (7)

Tool name	Description
puppeteer_navigate	Navigate to a URL in the browser
puppeteer_screenshot	Take a screenshot of the current page
puppeteer_click	Click an element on the page
puppeteer_fill	Fill a form field with text
puppeteer_evaluate	Execute JavaScript in the browser
puppeteer_select	Select an option from a dropdown
puppeteer_hover	Hover over an element

Example Prompts

“Take a screenshot of example.com”
“Navigate to my app and click the login button”
“Scrape the product names and prices from this e-commerce page”
“Fill in the contact form on my website with test data”
“Get the text content of all h2 headings on this page”

Pros & Cons

Pros

Full browser control — not just fetching HTML
Can interact with JavaScript-heavy sites
Takes screenshots for visual verification
Official Anthropic server

Cons

Requires Node.js and Chrome/Chromium installed
Slower than a simple HTTP fetch for static pages
Not suitable for headless server environments

Who Is This For?

·Developers automating browser-based workflows
·QA engineers who want AI-assisted UI testing
·Data engineers scraping JavaScript-rendered content

When to use: Use Puppeteer when you need to scrape or interact with pages that require JavaScript execution, login, or user interaction. For static HTML pages, the Fetch server is faster and simpler.