Write Kubernetes Manifests from a Plain-English Description

Kubernetes manifests are full of boilerplate that's easy to get subtly wrong — label selectors that don't match, readiness probes with the wrong timing, missing resource limits. Describe what you need and let Claude handle the YAML.

Write Kubernetes manifests for a Node.js API service:
- Deployment with 3 replicas, rolling update strategy
- Image: my-registry/api:latest, port 3000
- Resource limits: 512Mi memory, 500m CPU
- Readiness probe on GET /health with a 5s initial delay
- ClusterIP Service exposing port 80 → container port 3000
- ConfigMap for non-secret environment variables

Claude will produce well-structured YAML with proper labels, selector consistency, and sensible defaults — the kind of boilerplate that's easy to get slightly wrong by hand.

spec:
  replicas: 3
  strategy:
    type: RollingUpdate
  template:
    spec:
      containers:
        - name: api
          resources:
            limits:
              memory: "512Mi"
              cpu: "500m"
          readinessProbe:
            httpGet:
              path: /health
              port: 3000
            initialDelaySeconds: 5

Ask it to add a HorizontalPodAutoscaler, Ingress, or NetworkPolicy as follow-up requests — each one builds on the labels and names already defined, so Claude keeps everything consistent.

Describing intent in plain English and letting Claude handle the YAML syntax beats the Kubernetes docs every time.