ERD_INDEXERS_CHECKLIST.md still leaves the health and notifications parity slice unchecked.
Operators already have connectivity rollups and reputation summaries, but they still lack a direct read path for raw indexer_health_event rows defined by ERD_INDEXERS.md.
The next efficient step is to expose recent health events end-to-end so the existing /indexers console can show failure detail and conflict timing without introducing a larger notification system yet.
Add stored procedures indexer_health_event_list_v1(...) and stable wrapper indexer_health_event_list(...) to read recent events for one indexer instance with actor validation and bounded limits.
Extend the data, app, and API layers with typed health-event list reads and a new GET /v1/indexers/instances/{indexer_instance_public_id}/health-events route.
Extend the indexer admin UI with a health-event limit field, fetch action, and rendered drill-down cards under the connectivity section.
Keep notification delivery out of scope for this slice; the checklist item remains open until delivery hooks exist.