rkvalandas/browser_agent ? reverse-engineered prompt

Build me a Python app that lets me control a web browser with plain English commands.

I want to be able to start it from the command line, check that everything is working, then give it a task like opening a site, searching for something, filling out a form, clicking buttons, scrolling, and moving through pages on its own. It should be smart enough to understand what is on the page, handle normal dynamic websites, and ask me a question if it needs missing information before continuing.

Please make it work with Chrome, and let it either connect to an existing browser or open a new one. I also want a simple way to set up my API keys and choose between supported AI providers. If you need to look up current docs online for the browser automation parts, go ahead. Keep the experience easy to use, with clear commands for running the agent and checking status.