# AgentDeskAI/browser-tools-mcp

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

Repository: https://github.com/AgentDeskAI/browser-tools-mcp
Canonical: https://ross.abutalabs.com/products/browser-tools-mcp
Homepage: https://agentdesk.ai/
Language: TypeScript
License: MIT
License Family: permissive
Topics: ai, anthropic, cursor, mcp, mcp-server
Last push: 2026-08-12T16:48:04+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 97, release rhythm 97, longevity 40
- inputs: {"age_days": 569, "days_push": 21, "days_rel": 21, "gap_med": 11.5, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7300, forks 528 (observed 2026-08-28T04:09:58.111084+00:00)

## What it is
An MCP server plus Chrome extension that streams console logs, network activity, screenshots, and Lighthouse audits from a user's real, logged-in Chrome session to AI coding agents like Cursor, Claude Code, and Windsurf. Unlike CDP-based automation servers, it attaches to the developer's actual browser session via a DevTools extension, preserving auth state.

## Use cases
- let my AI agent read browser console errors while I debug
- capture network requests from my logged-in Chrome session for Cursor
- run accessibility and performance audits on the page I'm viewing via Claude Code
- debug my app without recreating auth state in a throwaway browser profile
- stream browser screenshots to an MCP-compatible IDE
- check console logs from my real browser session in Windsurf or Cline

## When to choose
- you want your AI coding agent to see what's happening in your real, logged-in Chrome session
- you need Lighthouse-grade performance, accessibility, or SEO data fed to an agent
- you use Cursor, Claude Code, Windsurf, Cline, Zed, or another MCP client for web debugging
- Chrome 136+ blocks remote debugging on your default profile and you want to avoid throwaway profiles

## When to avoid
- you need to drive a fresh automated browser for writing end-to-end tests (use Playwright MCP or Chrome DevTools MCP)
- you don't use any MCP-compatible client or AI coding agent
- you need headless or CI browser automation
- you can't run Node 22.19 or newer

## Facets
- artifact type: service
- maturity: active
- function: mcp, monitoring, logging, developer-tools
- domain: developer-tools, web-development, artificial-intelligence
- platform: browser, cross-platform, cli
- tags: mcp-server, browser-debugging, chrome-extension, cursor, claude-code, lighthouse-audits, console-logs, network-monitoring, debugging, ai-agents, nodejs

## Member repositories
- AgentDeskAI/browser-tools-mcp (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:58.111084+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:38:42.012315+00:00, confidence not recorded.
  - readme: https://github.com/AgentDeskAI/browser-tools-mcp (fetched 2026-08-28T04:09:58.111084+00:00, sha 1cfba86cb55d)
  - homepage: https://agentdesk.ai/ (fetched 2026-08-29T08:34:19.378334+00:00, sha 30e1dc61957c)
- Data as of 2026-08-30T08:39:29.467469+00:00.