A public status view for runtime health, provider freshness and known service degradation. A healthy web shell does not by itself mean every scientific provider is current, so provider state is reported separately.
Application health and provider health are reported separately
Public status loads the API state and provider list and checks which upstream sources have recently succeeded. The web application can be available while one scientific provider is failing or delayed. Service health is therefore not reduced to one green or red indicator.
Each provider has its own last success and failure
Provider cards show provider_id, current state, last_success_at and last_failure_at when present. Provider counts and provider_failures give a quick operational overview but do not replace the freshness policy for a specific metric. A provider may respond technically while one product is still too old to be considered Live.
Technical status is not a guarantee of complete data coverage
No provider failure does not mean every optional layer is available or every record has GOOD quality. First-party Schumann, weather, lightning or other inputs may remain UNAVAILABLE under their own contracts. Service Status helps diagnose operations, while Data Sources and methodology explain what the data mean.