Fleet operator + AWS-native compliance lens — the next layer after v1.1's K8s + Helm completeness.
Three themes:
- Cluster shell with single-log audit. Periscope-proxied kubectl, so every shell-issued API call lands in the existing structured audit log alongside UI-driven actions. Single-log-or-nothing.
- AWS compliance lens. Pod Identity Association viewer (per-pod IAM role) + CloudTrail pod-correlation table — every AWS API call a pod made in its lifetime, joined by pod session ARN. The killer feature for compliance teams chasing "what did this pod do to AWS?" across two log sources.
- Contextual metrics. Prometheus + CloudWatch Container Insights backends rendering CPU / memory sparklines on Pod / Deployment / Service / Node detail pages. NOT a metrics explorer; "open in Grafana / CloudWatch" deep links cover the deeper drill.
Also rolling forward from v1.1's "deferred" list: Helm chart-repo registry, Helm sub-chart dependencies.
Out of scope (v1.3+): MCP integration (waiting on AWS EKS MCP to mature), Datadog metrics, EKS Upgrade Insights, EKS Add-ons viewer.
Issues open after v1.1 lock — the milestone description tracks shape; sub-tasks land when v1.1 ships.