← Back to directory
P

Puppeteer MCP Server

Official
Archived
Headless browser control & screenshots
GitHub source repository ↗
★ 294 Stars Category · Browser Popular Source revision 9be4674d1ddf
62FMRS · C
Reliability
14/20
Security and permissions
11/20
Maintenance
10/20
Documentation
14/20
Setup experience
13/20

Browser automation is powerful and setup is direct, but web side effects and session credentials widen the risk surface; Chromium launch and MCP handshake evidence are not yet recorded.

Read the FMRS scoring method →

This server drives Chromium to navigate pages, fill forms, click elements, and capture screenshots. It lets models inspect and interact with rendered web pages for browser automation, testing, and data collection.

Tools

navigate
Navigate to a specified URL
screenshot
Capture a screenshot of the current page
click
Click an element on the page
fill
Fill a form input
evaluate
Run a custom script in the page

Setup

Start the server with npx. The required Chromium build is downloaded automatically on first run.

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

Fit and risk

Best for

  • Testing and collection tasks that need an assistant to inspect rendered pages
  • Development teams automating screenshots, form filling, or browser regression checks

Not for

  • Environments that prohibit automated external browsing or web-account credentials
  • Production-scale scraping that requires advanced anti-bot handling and distributed execution

Required permissions

  • Permission to launch a local Chromium process
  • Network access to target websites and, for authenticated flows, a suitable test account

Risks and side effects

  • Automated clicks and form fills can trigger real submissions or external side effects
  • Screenshots, page content, and browser sessions may expose sensitive information

Troubleshooting

  1. Confirm Chromium downloaded successfully and is executable
  2. Use the correct headless settings in environments without a desktop
  3. Check target-site connectivity, proxy settings, and certificates

Use cases

Run regression checks on important website flows
Capture scheduled screenshots for page archives
Collect data from pages that require JavaScript rendering

Supported clients

Claude DesktopFull support
CursorFull support
ClineFull support