pearlhimaliya03094299179-ux/wifi-densepose — reverse-engineered prompt

Build me a production ready Python app that uses WiFi CSI data from standard routers to estimate human body pose without cameras. It should process incoming router data, clean the signal, run a pose model, track multiple people, and stream live pose results through an API and websocket so another app can show it in real time.

Include simple setup instructions, sample config, a mock data mode for testing without hardware, authentication, rate limiting, logs, health checks, and tests. Add analytics for things like fall detection, activity recognition, and occupancy counts.

Make it easy to run locally with Docker, and leave clear notes for setting up WiFi hardware and calibration. If you need exact FastAPI or CSI details, look up current docs online, but keep the finished project practical and runnable.