How to set up Browser Use CLI (automate any browser task)

a command-line tool that controls your real chrome — logins, cookies, sessions intact — at ~50ms speed. install, connect it to claude code, and automate the boring stuff.

You can automate any browser task from your terminal.

What is Browser Use CLI?

A command-line tool that controls your browser at ~50ms speed. It uses your real Chrome profile (logins, cookies, sessions). It can navigate, click, fill forms, upload files — all from your terminal.

The key benefit: No more logging in again. No sandboxed browser. YOUR browser, automated.

Step 1 — Install it (30 seconds)

curl -fsSL https://browser-use.com/cli/install.sh | bash

Requires Python 3.11+. Run browser-use doctor to verify everything works.

Step 2 — Connect it to Claude Code (optional)

npx skills add https://github.com/browser-use/browser-use --skill browser-use

This lets your Claude Code agent use the browser CLI as a skill — so your agent can browse the web for you.

Step 3 — Run your first command

Basic commands:

browser-use open https://yoursite.com — opens a page
browser-use click 3 — clicks element #3
browser-use type "hello" — types text
browser-use screenshot — captures what you see
browser-use state — shows all interactive elements

It keeps a persistent session so you don't restart the browser every time.

What I use it for

I pointed it at my shipping website and automated:

Go through all pending orders
Approve each one
Schedule the pickup
Generate shipping labels

15-20 min of daily manual work → done by AI.

Standout features

Named sessions, JS execution, cookie import/export, and cloud browser support.

Full reference: docs.browser-use.com/open-source/browser-use-cli