fatihmakes/mark-lv ? reverse-engineered prompt

Build me a cross platform voice assistant app that feels like a real JARVIS style companion for Windows, macOS, and Linux.

I want it to listen with push to talk and wake word, speak back naturally, show a live animated avatar and waveform on screen, and let me control my computer by voice like opening apps, moving windows, managing volume, brightness, WiFi, browser tabs, files, and basic power actions. It should also be able to see my screen or webcam when needed, remember useful things about me across sessions, and show a memory panel where I can review or delete what it knows.

Please include a simple dashboard or HUD UI, support for live voice and voice changes, a way to play YouTube or local videos inside the assistant screen, and a plugin system so I can drop in new skills later. If you need to check current docs or libraries to make it work smoothly, go ahead and look them up online.