Reverse engineered prompt

GitHub

Build me a simple Python tool that can look at a Rubik’s Cube in either a single photo, a folder of side photos, or a live webcam feed, find the cube face in view, and print the RGB value for every sticker square it detects. It should work for 2x2x2, 3x3x3, 4x4x4, 5x5x5, and 6x6x6 cubes, and ideally larger ones too if the sticker sizes are consistent.

For webcam mode, I want to be able to scan each side by pressing the spacebar, reset with the r key, and have the app guide me through scanning in F R B L U D order. After all sides are scanned, if it is a 2x2x2, 3x3x3, or 4x4x4, show the solve steps on screen. For a single image mode, it is fine if it opens debug views while figuring out the square locations. Also support reading a directory of images named by face.

Please make it runnable from the command line and look up any current docs online if you need to.

Have a live product UI? Try website reverse