puppeteer-real-browser-mcp-server

Puppeteer Real Browser MCP Server

Puppeteer Real Browser connects AI assistants to a real browser for navigation, form entry, and content extraction. It serves Claude and Cursor users automating website interactions through MCP.

Server facts

puppeteer-real-browser-mcp-server

Tech

Runtime / language
node
Transports
stdio
Remote available
No
Authentication
Not specified

Activity

GitHub stars
26
Last commit
Data as of

Overview

Puppeteer Real Browser MCP Server exposes browser interactions to AI assistants using ZFC Digital’s puppeteer-real-browser package. A session begins with browser_init, which includes Chrome path detection and anti-detection features. Assistants can then navigate pages, click elements, enter text, and wait for conditions before continuing. For locating controls, find_selector identifies a CSS selector from an element’s text. Content workflows include retrieving HTML or text with get_content and exporting extracted material to a formatted Markdown file with save_content_as_markdown. The browser_close tool ends the browser instance.

Additional tools provide scrolling with randomized timing and an attempt to solve CAPTCHAs where supported; CAPTCHA success is not guaranteed. Browser initialization also accepts options such as headless operation and proxy configuration. The Node-based server communicates over stdio, with setup guidance for Claude Desktop, Claude Code CLI, and Cursor. It requires Node.js 18 or later and an installed Chrome or Chromium browser. The project is under active maintenance, and its documentation warns that features may remain incomplete or change without notice.

puppeteer-real-browser-mcp-server setup notes

  • puppeteer-real-browser-mcp-server records stdio only; confirm a network transport upstream before planning remote access.
  • puppeteer-real-browser-mcp-server has no recorded license; confirm reuse terms in its repository.

Related MCP servers