I’ve been experimenting with ways to get Cursor to control a browser for automating tasks (clicking, typing, scrolling, etc.). The idea is to have Cursor interact with a library that can parse the DOM / take screenshots and then feed that into Cursor and generate an action. The library can then execute the action on the browser.
My project is called Cesail and I put it on Github if anyone wants to check it out. I’m curious if anyone here has tried something similar with Cursor, or has ideas on how to improve my tool.