Operator daily-driver layer — fill the everyday-pane features still missing after v1.1, with GPU + AI workload visibility as the headline.
gpu + ai workload visibility
— pod ↔ gpu mapping layer (podresources gRPC + dcgm + dra + cost), idle gpu finder, dcgm reconciler.
8 issuesoperator shells
— in-browser cluster shell with single-row audit + rbac pre-flight; ssm-into-eks-nodes with per-user impersonation.
2 issuesrelated-resources foundation
— rfc 0005 phase 0 — relationshipindex skeleton + pod ownerreferences extractor + <RelatedList> on the pod detail pane.
1 issuehelm private oci
— pull from private ECR / GHCR / artifact registry via pod identity / IRSA — no static creds.
1 issueother
— issues not yet themed
1 issue