retrovibesgenz-blip/AgentLab ? reverse-engineered prompt
Reverse engineered prompt
Build me a Windows desktop AI assistant app that feels like my own Jarvis.
I want to talk to it or type to it, and have it actually use my computer for me, opening apps, searching the web, playing music, clicking buttons, typing text, scrolling, and handling shortcuts like a real person would. It should show a chat panel and an on screen cursor overlay so I can watch what it is doing, and it should be able to handle more than one agent at once with separate cursors and simple pause, resume, and stop controls.
Please make voice chat work too, with speech in and speech out, and use an ElevenLabs voice if I add my API key. I also want the option to choose between a cloud vision model, a Kaggle hosted model, or a local model running in the app, with a settings screen or easy config for switching providers. If you need current docs for any APIs, look them up online.
Are you gonna build this?
make sure you review the code using arcumet