chrome-cdp-skill

Why we included this project

Instead of spinning up a separate browser for an agent to drive, this skill connects to the Chrome session you already have open, so the agent can read pages you are logged into and work with the tabs you are actively using. It offers a small set of commands for listing tabs, taking screenshots, pulling the accessibility tree or HTML, clicking, typing, navigating, and evaluating JavaScript in the live page. Setup is minimal: install it as a pi skill, or drop the skill directory into Amp, Claude Code, or Cursor, with Node.js as the only runtime dependency and a single toggle in Chrome's remote debugging settings. For developers who want an agent to handle real authenticated workflows in the browser they already trust, this is a direct and practical solution.