A9T9/RPA — reverse-engineered prompt

Build me an open source browser automation extension like Ui.Vision RPA. I want a normal user to be able to record a web task, save it as a macro, replay it later, and use it for things like filling forms, checking websites, scraping simple data, and running data driven tests.

It should work in Chrome, Edge, and Firefox if possible. Include a simple Selenium IDE style import and export so people can bring in existing scripts. Add visual automation helpers too, like OCR and computer vision based commands for finding text or images on the screen, not only regular web selectors.

Please make the project easy to run locally, with clear install, dev, and build commands. I’d like the final build folders ready for Chrome and Firefox, plus a short README that explains how to install the extension manually and how to create and run a first macro. Look up current browser extension docs online if you need to.