ritesh1307be22-lang/Zarvis- — reverse-engineered prompt
Reverse engineered prompt
Build me a Python Streamlit app called Zarvis, a simple AI virtual assistant that I can talk to with my microphone. It should listen for the wake word “Jarvis”, understand voice commands, speak responses back, and use a local Phi 3 model through Ollama for private AI answers.
I want it to feel like a personal desktop helper. It should open common websites like Google, YouTube, and LinkedIn, send WhatsApp messages, get weather and latest news, take screenshots, capture photos with the camera, show CPU and battery status, and include a focus mode or workspace cleanup option. Please also show a clean Streamlit interface with buttons, status messages, and a session history of what I asked and what Zarvis did.
Make it easy to run with clear setup steps, required packages, and sensible error messages if the microphone, camera, or Ollama are not working. Look up current docs online if needed.
Want more depth? Deep Reverse