zestones/Aria — reverse-engineered prompt
Reverse engineered prompt
Build me a local demo app for factory maintenance called ARIA. I want something that feels usable by an operator, not just a backend. The flow should be, upload a machine manual PDF, ask a few simple operator knowledge questions, and turn that into a living knowledge base for the machine. Then keep watching simulated machine signals, logbook notes, shift handover notes, failure history, and basic KPIs like OEE, MTBF, and MTTR.
When the system notices drift or an anomaly, it should surface a warning in the UI, do a deeper diagnosis, explain the likely root cause in plain English, and create a practical work order with actions, parts, and a suggested intervention window. I also want a chat style Q and A area where an operator can ask what is going on and get answers grounded in the machine history.
Please make it easy to run locally with Docker, seeded demo users, and a fast demo mode with simulated failure scenarios. Look up current docs online if you need to.
Want more depth? Deep Reverse