Chrome DevTools for agents (formerly Chrome DevTools MCP, npm package chrome-devtools-mcp) is an MCP server from the Chrome DevTools team. It lets a coding agent such as Antigravity, Claude, Cursor, Copilot, Codex or Gemini CLI control and inspect a live Chrome browser. A CLI is also provided for use without MCP.
The agent can record performance traces, analyze network requests, read console messages with source-mapped stack traces, take screenshots, fill forms, click, emulate devices and locations, and run Lighthouse audits. It uses Puppeteer for automation. It can start its own Chrome or connect to a Chrome session you already have open.
It runs locally with Node.js (LTS), npm and current stable Chrome. It is free and open source under the Apache-2.0 licence. Google collects usage statistics by default; the --no-usage-statistics flag turns this off.
Developers who use AI coding agents to build websites and want the agent to test, debug and verify its code in a real Chrome browser. Also for makers of agent tools who want to add a browser subagent.
Records performance traces and runs Lighthouse audits with source-mapped console messages from Chrome DevTools; it also ships a CLI for use without MCP.
Runs on your own computer and can connect to the Chrome session you already have open; Cloudflare's Browser Run server is hosted and needs a Cloudflare account.
Free under Apache 2.0 and needs no account: it inspects the live page while you develop, before any error reaches a monitoring service.
MCP server from Microsoft that lets AI models control a browser with Playwright
Supports Chrome, Firefox, WebKit and Edge and can run from a Docker image; Chrome DevTools for agents works with Chrome only.
Sentry's official MCP server that connects coding agents to Sentry
Reads errors, traces, logs and replays already collected in Sentry from production, and can ask Seer for a root cause; Chrome DevTools for agents inspects one live browser.
讨论
登录以参与讨论