What problem does it solve? Kubernetes changes often fail in production because manifests are reviewed for syntax rather than operational behavior. This Skill analyzes Deployments, StatefulSets, RBAC, NetworkPolicies, probes, and storage configuration against real rollout, scheduling, and failure scenarios so issues are caught before they reach a cluster. ## Core Features & Use Cases - Workload and rollout review: Evaluates Deployment/StatefulSet/DaemonSet update strategies, probe correctness, and resource requests against realistic startup and steady-state behavior. - Security and tenancy hardening: Checks RBAC bindings, service accounts, secret mounting, namespace isolation, and NetworkPolicy rules against least-privilege and intended traffic paths. - Failure-path troubleshooting: Diagnoses crash-looping pods, autoscaling thrash, and storage durability risks during rescheduling, separating confirmed evidence from assumptions about live cluster state. - Use Case: Paste a Deployment manifest and its logs for a crash-looping pod, and receive the root-cause hypothesis, the smallest safe fix, and a validation plan covering normal, failure, and rollback paths. ## Quick Start Ask the agent to review your Kubernetes manifest or Helm values and explain why your pod is crash-looping, including the smallest safe fix and rollback plan.
