With an Aqara setup spread across M3, M100, and M200, I’ve found it useful to keep one simple, noncritical “canary” device on each hub—ideally something easy to trigger repeatedly, such as a plug, bulb, or basic contact sensor.
After a hub firmware, Matter bridge, or Apple Home update, I test those canaries before testing complex scenes:
- Verify the device responds in Aqara Home.
- Verify the same accessory or bridged endpoint responds in Apple Home.
- Run one small automation tied only to that hub.
- Record latency and whether the failure appears on one hub, one exposure path, or everywhere.
If only one canary fails, the search stays local to that hub or bridge. If all three fail only in Apple Home, the controller or Matter/HomeKit layer becomes the stronger suspect. It is much faster than using a large scene as the first test, because a scene can hide which dependency actually broke.
Do you keep a dedicated test accessory per hub, or reuse a production device when checking hub health?