SHIP status and availability page. Each team could expose metrics, health endpoints, etc.. and we could surface them all on a status webpage backed by a REST API. One place to go to see if payloads are flowing, build clusters are up, prow components are working, accepted payloads have happened in the last day (maybe?). We could hook into this with an MCP server to utilize it from AI. It would not only report the current status, but also list historical outages so that an AI (or humans) could tell if a build cluster was having trouble at a certain time.

This requires cross-functional effort, as other teams within SHIP would have to add their own configuration for components they would like to report the status of.

The highest leverage areas would be build clusters and prow availability.