[gianlucam76]

An EventTrigger can match hundreds of resources per cluster. /event returned all of them for every matching cluster in a single response, which made the payload large and the dashboard page slow. This change returns only the number of matching resources per cluster, and adds a paginated endpoint to list them on demand.

Changes

• /event: each entry in clusterEventMatches now has matchingResourcesCount instead of the resources array. This is a breaking change to the response. The dashboard is the only consumer: mcp-server reads EventReports directly.
• New GET /eventResources?event_name=&namespace=&name=&type=<capi|sveltos>:
  • Returns totalResources and one page of resources, ordered by group, kind, namespace and name.
  • Supports limit and skip.
  • Supports filters resource_kind, resource_namespace and resource_name, as a case-insensitive substring match. totalResources is counted after filtering and before pagination.
  • The user needs permission to get both the EventTrigger and the cluster. The response is 400 for a bad request, 404 for an unknown EventTrigger and 401 without permission. A cluster with no EventReport returns an empty list.
• README updated for both endpoints.