koetjeengdjalanan/AFIRA — reverse-engineered prompt

Build me a Python automation service called AFIRA that runs all the time and pulls network information from HPE Aruba Central, then stores it in InfluxDB so I can make Grafana dashboards from it.

It should use Aruba Central OAuth credentials from a separate creds file, read normal settings from an env file, and repeat on a schedule like every 5 minutes. Each run should collect site health, device inventory and status, WLAN info, client data, locations, web app metrics, and access point or switch hardware stats like CPU, memory, power, and switch details. If something fails, log it clearly and keep trying on the next cycle instead of crashing.

Please include a simple Docker setup with InfluxDB and Grafana, plus a way to run the collector locally or in a container. Make sure secrets don’t get baked into the image. Add useful logging to a file and optionally into InfluxDB, graceful shutdown, basic tests, and clear setup instructions for editing env values and creds.