maintenance in HA is usually reactive—we fix things only when they break. to make it proactive and a bit more fun, i built HAGHS (HA global health score).
it’s a custom integration providing a score from 0 to 100 based on:
40% hardware: cpu tiers, ram, and disk pressure
60% hygiene: “zombie” entities, database bloat, and version lag
i think spook is great for specific repairs, while haghs is a high-level benchmark for overall system health. they complement each other: spook is the scalpel, haghs the heart rate monitor. for me, using both is the ultimate way to keep your instance clean.