AIwithhassan/likhari-ai ? reverse-engineered prompt
Reverse engineered prompt
Build me a small desktop app for voice dictation that works on macOS, Windows, and Linux.
I want to hold the right Alt key to talk, see a live transcript while I’m speaking, then let go and have the app clean up the text and paste it into whatever app I’m using without taking focus away. It should use local speech recognition first, with an optional cloud or local AI cleanup step to make the text read better, fix punctuation, and remove filler words. If no AI cleanup is available, it should still work with simple offline cleanup.
Please include a tray or menu bar icon, a quick mute option, and a settings window where I can choose the microphone, transcription model, cleanup behavior, and a personal dictionary for custom word replacements. It should feel lightweight, launch on login, and handle common audio problems like disconnected mics or silence gracefully. If you need to look up current docs online, go ahead.
Are you gonna build this?
make sure you review the code using coderabbit