samShojaiefar/Voca ? reverse-engineered prompt

Build me a small desktop voice dictation app with a tiny notch style overlay that stays out of the way.

When I press and hold a key, it should start listening, show a clear visual indicator, and type my spoken words into whatever app I’m using. On first launch, I want a simple setup flow that walks me through microphone and accessibility permissions, choosing the push to talk key, picking a speech model, doing a quick test, and setting a primary language plus an optional second language.

It should support automatic language detection, let me switch between my two languages with Ctrl Shift L, and understand emoji names so it can insert them too. Keep the UI clean and minimal, with a nice Apple inspired feel. If you need to, look up current docs online for anything that changes with the tools or APIs.