kaushaljindall/AirInk — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a Python webcam app called AirInk where I can draw letters in the air with my hand and have them turn into typed words on screen.

I want it to use real time hand tracking from the webcam. When I hold up only my index finger, it should draw on a canvas. When I hold up index and middle fingers, it should let me choose a brush from a top color bar. Include purple, blue, green, and an eraser. When I open all five fingers, it should recognize the letter I just drew, flash the detected letter big on the screen, add it to a word bar at the bottom, and then clear the canvas so I can draw the next letter.

Also add keyboard controls for space, backspace, clear, and quit. Show an FPS counter too. Use OCR for single letter recognition and make the setup work on Windows with Python. Keep the code simple enough that I can run it locally, and include any install notes needed. Look up current docs online if you need to.

Want more depth? Deep Reverse