BrowserMCP

Name: BrowserMCP
Author: BrowserMCP

by BrowserMCP

Control your real browser with AI using your existing logged-in sessions

browser Node.js Beginner Self-hostable No API key

⭐ 5.9k stars 📅 Updated: 1mo ago

View on GitHub

Description

An MCP server paired with a Chrome extension that enables AI applications to automate your actual browser — not a headless instance. Unlike traditional browser automation, BrowserMCP uses your real browser profile with all existing logged-in sessions, cookies, and extensions intact. This means the AI can interact with authenticated services (Gmail, GitHub, internal tools) without needing credentials. Derived from Microsoft's Playwright MCP but redesigned for user-browser automation. Runs locally with no data leaving your device, and uses your real browser fingerprint to bypass basic bot detection.

✅ Best for

Users who want AI assistants to interact with their authenticated web services directly

⏭️ Skip if

You need headless browser automation or cross-browser support

💡 Use cases

Automating tasks in authenticated web services without sharing credentials
AI-driven browser automation using your existing logged-in sessions
Interacting with internal tools and dashboards through natural language
Quick web tasks (form filling, data extraction) from AI assistants

👍 Pros

✓ Uses your real browser — all sessions, cookies, and extensions available
✓ No credentials needed for authenticated services
✓ Local execution — browser activity stays on your device
✓ Real browser fingerprint avoids basic bot detection

👎 Cons

✗ Requires Chrome extension installation alongside the MCP server
✗ Cannot be built independently from the monorepo (dependency issue)
✗ Limited to Chrome/Chromium browsers only

🔧 Exposed tools (6 tools)

Tool	Category	Description
browser_click	interaction	Click on page elements in the real browser
browser_type	interaction	Type text into input fields
browser_navigate	navigation	Navigate to a URL in the user's browser
browser_scroll	navigation	Scroll the page up or down
browser_snapshot	observation	Capture accessibility snapshot of the current page
browser_screenshot	observation	Take a screenshot of the current page

⚡ Installation

Prerequisites:

• Node.js v18+
• chrome-extension

Check Claude Code documentation to configure this MCP server.

💡 Tips & tricks

Install the Chrome extension first from browsermcp.io, then configure the MCP server in your AI client. Documentation is at docs.browsermcp.io.

🔄 Alternatives

Playwright MCP

Browser automation for LLMs via accessibility snapshots, no vision models needed

Similar functionality

browser-use

MCP server — editorial content pending