Browser Use
Persistent browser automation CLI for navigating websites, filling forms, taking screenshots, extracting data, and running autonomous cloud browser agents — all from your AI agent.
# Browser Use Persistent browser automation CLI for navigating websites, filling forms, taking screenshots, extracting data, and running autonomous cloud browser agents — all from your AI agent. Browser Use provides fast, persistent browser automation through a CLI. Your AI agent can navigate websites, inspect page elements, click buttons, fill forms, take screenshots, extract data, execute JavaScript, manage cookies, and run autonomous browser agents in the cloud. Browser sessions persist across commands, enabling complex multi-step workflows like authenticated browsing, web testing, and data extraction. ## How It Works The skill operates through the browser-use CLI with a navigate-inspect-interact-verify workflow. First, open a URL to start or reuse a browser session. Then run state to get a list of clickable elements with indices. Interact with those elements by index (click, type, select). Verify results with state or screenshot . The browser stays open between commands, maintaining state across interactions. Three browser modes are available: headless Chromium (fast, isolated), real Chrome (with your login sessions and profiles), and remote cloud browsers (with proxy support and parallel agent execution). ## Key Features - **Persistent Sessions**: Browser stays open across commands, maintaining cookies, state, and navigation history - **Three Browser Modes**: Headless Chromium for speed, real Chrome with profile support for authenticated browsing, and cloud remote browsers for parallel tasks - **Element Interaction by Index**: Inspect page state to get clickable elements with indices, then click, type, hover, double-click, or right-click by index - **Screenshot Capture**: Take full-page or viewport screenshots as base64 or saved to file - **JavaScript Execution**: Run arbitrary JS in the browser context and return results - **Cookie Management**: Get, set, clear, import, and export cookies with full attribute control - **Profile Syncing**: Sync cookies from local Chrome profiles to cloud profiles for authenticated remote browsing - **Cloud Agent Tasks**: Launch autonomous browser agents in the cloud that run in parallel with configurable LLM models - **Tunnel Support**: Expose local dev servers to cloud browsers via Cloudflare tunnels - **Python Scripting**: Persistent Python session with access to the browser object for complex automation ## Requirements - **browser-use CLI**: Must be installed and configured. Run browser-use doctor to verify the installation ## Use Cases - **Web Testing**: Automate UI testing by navigating pages, filling forms, and verifying results with screenshots - **Data Extraction**: Scrape structured data from websites using element inspection and JavaScript execution - **Form Filling**: Automate repetitive form submissions across web applications - **Authenticated Browsing**: Access sites you are logged into by using your Chrome profile with saved sessions - **Parallel Research**: Launch multiple cloud browser agents simultaneously to research different topics - **Local Dev Testing**: Expose localhost dev servers via tunnels and test them in cloud browsers ## Installation Install via: `npx clawhub@latest install browser-use`