Agent Skills for Claude Code | Kubernetes Specialist


Domain	Infrastructure & Cloud
Role	specialist
Scope	infrastructure
Output	manifests

Triggers: Kubernetes, K8s, kubectl, Helm, container orchestration, pod deployment, RBAC, NetworkPolicy, Ingress, StatefulSet, Operator, CRD, CustomResourceDefinition, ArgoCD, Flux, GitOps, Istio, Linkerd, service mesh, multi-cluster, cost optimization, VPA, spot instances

Related Skills: DevOps Engineer · Cloud Architect · SRE Engineer

When to Use This Skill

Deploying workloads (Deployments, StatefulSets, DaemonSets, Jobs)
Configuring networking (Services, Ingress, NetworkPolicies)
Managing configuration (ConfigMaps, Secrets, environment variables)
Setting up persistent storage (PV, PVC, StorageClasses)
Creating Helm charts for application packaging
Troubleshooting cluster and workload issues
Implementing security best practices

Core Workflow

Analyze requirements — Understand workload characteristics, scaling needs, security requirements
Design architecture — Choose workload types, networking patterns, storage solutions
Implement manifests — Create declarative YAML with proper resource limits, health checks
Secure — Apply RBAC, NetworkPolicies, Pod Security Standards, least privilege
Validate — Run kubectl rollout status, kubectl get pods -w, and kubectl describe pod <name> to confirm health; roll back with kubectl rollout undo if needed

Reference Guide

Load detailed guidance based on context:

Topic	Reference	Load When
Workloads	references/workloads.md	Deployments, StatefulSets, DaemonSets, Jobs, CronJobs
Networking	references/networking.md	Services, Ingress, NetworkPolicies, DNS
Configuration	references/configuration.md	ConfigMaps, Secrets, environment variables
Storage	references/storage.md	PV, PVC, StorageClasses, CSI drivers
Helm Charts	references/helm-charts.md	Chart structure, values, templates, hooks, testing, repositories
Troubleshooting	references/troubleshooting.md	kubectl debug, logs, events, common issues
Custom Operators	references/custom-operators.md	CRD, Operator SDK, controller-runtime, reconciliation
Service Mesh	references/service-mesh.md	Istio, Linkerd, traffic management, mTLS, canary
GitOps	references/gitops.md	ArgoCD, Flux, progressive delivery, sealed secrets
Cost Optimization	references/cost-optimization.md	VPA, HPA tuning, spot instances, quotas, right-sizing
Multi-Cluster	references/multi-cluster.md	Cluster API, federation, cross-cluster networking, DR

Constraints

MUST DO

Use declarative YAML manifests (avoid imperative kubectl commands)
Set resource requests and limits on all containers
Include liveness and readiness probes
Use secrets for sensitive data (never hardcode credentials)
Apply least privilege RBAC permissions
Implement NetworkPolicies for network segmentation
Use namespaces for logical isolation
Label resources consistently for organization
Document configuration decisions in annotations

MUST NOT DO

Deploy to production without resource limits
Store secrets in ConfigMaps or as plain environment variables
Use default ServiceAccount for application pods
Allow unrestricted network access (default allow-all)
Run containers as root without justification
Skip health checks (liveness/readiness probes)
Use latest tag for production images
Expose unnecessary ports or services

Common YAML Patterns

Deployment with resource limits, probes, and security context

apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-app
  namespace: my-namespace
  labels:
    app: my-app
    version: "1.2.3"
spec:
  replicas: 3
  selector:
    matchLabels:
      app: my-app
  template:
    metadata:
      labels:
        app: my-app
        version: "1.2.3"
    spec:
      serviceAccountName: my-app-sa   # never use default SA
      securityContext:
        runAsNonRoot: true
        runAsUser: 1000
        fsGroup: 2000
      containers:
        - name: my-app
          image: my-registry/my-app:1.2.3   # never use latest
          ports:
            - containerPort: 8080
          resources:
            requests:
              cpu: "100m"
              memory: "128Mi"
            limits:
              cpu: "500m"
              memory: "512Mi"
          livenessProbe:
            httpGet:
              path: /healthz
              port: 8080
            initialDelaySeconds: 15
            periodSeconds: 20
          readinessProbe:
            httpGet:
              path: /ready
              port: 8080
            initialDelaySeconds: 5
            periodSeconds: 10
          securityContext:
            allowPrivilegeEscalation: false
            readOnlyRootFilesystem: true
            capabilities:
              drop: ["ALL"]
          envFrom:
            - secretRef:
                name: my-app-secret   # pull credentials from Secret, not ConfigMap

Minimal RBAC (least privilege)

apiVersion: v1
kind: ServiceAccount
metadata:
  name: my-app-sa
  namespace: my-namespace
---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: my-app-role
  namespace: my-namespace
rules:
  - apiGroups: [""]
    resources: ["configmaps"]
    verbs: ["get", "list"]   # grant only what is needed
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: my-app-rolebinding
  namespace: my-namespace
subjects:
  - kind: ServiceAccount
    name: my-app-sa
    namespace: my-namespace
roleRef:
  kind: Role
  name: my-app-role
  apiGroup: rbac.authorization.k8s.io

NetworkPolicy (default-deny + explicit allow)

# Deny all ingress and egress by default
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: default-deny-all
  namespace: my-namespace
spec:
  podSelector: {}
  policyTypes: ["Ingress", "Egress"]
---
# Allow only specific traffic
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: allow-my-app
  namespace: my-namespace
spec:
  podSelector:
    matchLabels:
      app: my-app
  policyTypes: ["Ingress"]
  ingress:
    - from:
        - podSelector:
            matchLabels:
              app: frontend
      ports:
        - protocol: TCP
          port: 8080

Validation Commands

After deploying, verify health and security posture:

# Watch rollout complete
kubectl rollout status deployment/my-app -n my-namespace

# Stream pod events to catch crash loops or image pull errors
kubectl get pods -n my-namespace -w

# Inspect a specific pod for failures
kubectl describe pod <pod-name> -n my-namespace

# Check container logs
kubectl logs <pod-name> -n my-namespace --previous   # use --previous for crashed containers

# Verify resource usage vs. limits
kubectl top pods -n my-namespace

# Audit RBAC permissions for a service account
kubectl auth can-i --list --as=system:serviceaccount:my-namespace:my-app-sa

# Roll back a failed deployment
kubectl rollout undo deployment/my-app -n my-namespace

Output Templates

When implementing Kubernetes resources, provide:

Complete YAML manifests with proper structure
RBAC configuration if needed (ServiceAccount, Role, RoleBinding)
NetworkPolicy for network isolation
Brief explanation of design decisions and security considerations