agentdeskai/browser-tools-mcp

Monitor browser logs directly from Cursor and other MCP compatible IDEs.

View on GitHub ↗Jump to charts ↓Open shareable report

Summary Information

Updated 37 minutes ago
Added to GitGenius on September 9th, 2026
Created on February 10th, 2025
Open Issues & Pull Requests: 1 (+0)
GitHub issues: Enabled
Number of forks: 531
Total Stargazers: 7,315 (+1)
Total Subscribers: 31 (+0)

Repository Insights (GitGenius)

Median issue/PR response: 22.1 hours
Mean response time: 12.3 days
90th percentile: 36.9 days
Tracked items: 135

How this project is maintained

Around half of the issues opened in the past year never receive a reply. Only 9% of issues opened in the past year have been closed. Three people close 53% of everything that gets resolved.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Open issues: 1
New in 7 days: 0
Closed in 7 days: 0
Avg open age: 26 days
Stale 30+ days: 0
Stale 90+ days: 0

Recent activity

Opened in 7 days: 0
Closed in 7 days: 0
Comments in 7 days: 0
Events in 7 days: 0

Top labels

  • question (11)
  • help wanted (5)
  • enhancement (3)
  • bug (2)

Most active issues this week

No issue events were indexed in the last 7 days.

Detailed Description

BrowserTools MCP is an MCP server that streams browser diagnostics from a real Chrome session to AI coding agents in Cursor and other compatible IDEs.

The tool solves the problem of giving AI agents visibility into a live web application without requiring them to automate a fresh browser instance. Unlike Chrome DevTools Protocol-based servers that drive automated browsers, BrowserTools attaches to your existing Chrome session through a DevTools extension, keeping you logged in and on the page you were already viewing. It captures console output, network activity, screenshots, and Lighthouse audits—performance, accessibility, and SEO data—and relays them to the MCP client running in your editor.

BrowserTools suits developers who want their AI agent to debug or audit an application they are actively using, rather than setting up test automation. It eliminates the friction of recreating authentication state in a throwaway browser profile. The tool works with any MCP-compatible client including Cursor, Claude Code, Windsurf, Cline, Zed, and Gemini CLI. Installation requires Node 22.19 or newer and involves pointing your MCP client to the server and loading the Chrome extension from the repository; no separate connector process needs to be started.

The project underwent a significant rewrite in version 2.0, consolidating three processes into one, removing an unauthenticated local server, implementing credential scrubbing before data leaves the browser, and adding a test suite. A critical vulnerability in the prior version prompted the upgrade recommendation. The tool includes a diagnostic command to identify missing setup pieces and a verbose mode for observing capture in real time during installation verification.