Browser Harness | Browser Use CLI
Let your agentuse your browser.
Claude Code, Codex or Cursor clicks, types and reads pages in your own Chrome, logged in as you.
When a tool is missing,
the agent writes it.
No fixed tool list to outgrow. The agent adds what a site needs, uses it, and keeps it for next time.
- agent: wants to pick a date
- agent_helpers.py → helper missing
- agent writes itagent_helpers.py + pick_date()
- date picked
Questions
Is it my real browser?
Yes. The agent works in the Chrome you already use, with your tabs, cookies and logins, so it can do things on sites you are signed in to. Chrome asks once for permission.
Browser Harness or the Browser Use CLI?
Same engine. The Browser Use CLI (the browser-use command) is built on Browser Harness and ships with the rest of Browser Use. Browser Harness is the standalone open-source repo. Either one gives your agent your browser.
How do I know which tab it is using?
The tab your agent controls gets a 🐴 at the start of its title. It works in the background, so the tab you are looking at stays yours.
What can my agent do with it?
Anything you do in a browser: open pages, click, type, fill forms, read what is on screen, download and upload files. Ask in plain words, like "download my latest 20 X videos".
Which agents does it work with?
Claude Code, Codex, Cursor, Gemini CLI, Hermes Agent and OpenClaw. Anything that can run a shell command.
Can it run when my laptop is closed?
Yes, on Browser Use cloud browsers at $0.02 per browser-hour, with stealth, CAPTCHA solving and proxies. Run one per agent to do many tasks at once.
What does it cost?
Nothing. Browser Harness is open source under MIT, and using your own Chrome is free. Cloud browsers are $0.02 per browser-hour if you want them.