prateekhbtu/video-intel ? reverse-engineered prompt

Build me a working prototype of this edge to cloud video intelligence system.

I want an on premise camera agent that can read RTSP video, do motion filtering, run object detection and tracking, create simple zone events, and keep working even if the network drops. It should save data locally first, then sync events and telemetry to a cloud service when it can. Also add a control channel so the cloud can push down things like camera settings, thresholds, and retention rules.

On the cloud side, I want a small API that accepts events, keeps them idempotent, and supports a basic review or identity resolution flow. Please include the local and cloud data storage pieces, a simple verification script for invariants, and a lightweight dashboard if that fits. Keep it practical and make sure the project can run in a dev environment with simulated cameras too. If you need to check current docs for any library, feel free to look them up online.