After nearly losing a drive that SMART still rated "PASSED," I built SMART Sniffer - a HACS integration that actually watches the attributes that predict failure.
https://github.com/DAB-LABS/smart-sniffer
What it does:
- Monitors early warning SMART attributes (reallocated sectors, uncorrectable errors, wear leveling, pending sectors)
- Fires persistent HA notifications automatically when something starts climbing
- Multi-machine ~ lightweight Go agent on each machine, auto-discovered via mDNS
- One-liner install per machine, zero config in HA
- HAOS add-on available for monitoring the local host drive
What it doesn't do:
- No standalone web UI or database to maintain
- No MQTT, no InfluxDB, no Grafana
- Not a Scrutiny replacement ~ it's the "just tell me before it's too late" approach
Works on Linux, macOS, and Windows. Published ARM for RasPi. Needs testing.
HA INTEGRATION + AGENT (monitor any drive): https://github.com/DAB-LABS/smart-sniffer
HAOS APP (monitor HAOS drive + mock drives for testing): https://github.com/DAB-LABS/smart-sniffer-app