This repository implements a simple GUI intended to facilitate working with Browser Use in a Linux computer environment.
Browser Use is an incredible project for facilitating agentic control over a web browser.
This GUI simply provides a minimal Python wrapper over the app and has been validated to work on Open SUSE Tumbleweed Linux using KDE Plasma as the desktop environment.
Firstly, configure an API credential by copying .env.example to .env and providing your OpenAI API key.
Next, enter your prompt as the task description.
Hit the "Go" button!
Be prepared to be amazed.
Recommendation: Start with a very modest task to get a feel for how the program works.
The project's website has some suggestions around use-cases.
Using Browser Use entails what it says on the tin: giving an LLM not only programmatic access to your computer, but the ability to directly manipulate a GUI.
While its potential utility is vast, equally there is great potential for destruction! Close supervision is highly recommended, at least during the trial and error period.
To facilitate aborting the agent and for the amusement factor, a conspicuous stop button has been added, as well as a stop context menu in the system tray icon.
https://github.com/browser-use/browser-use