helm-k8s-manifest-reviewer

Purpose

Audit Kubernetes manifests (Deployments, Services, Ingress, HPAs) and Helm charts for production readiness, proper resource requests/limits, liveness/readiness probes, pod disruption budgets (PDB), and security context hardening.

When to Use

• Preparing a new service for deployment onto a Kubernetes cluster.
• Auditing Helm chart templates before release.
• Diagnosing CrashLoopBackOff, sudden pod evictions, or traffic drops during rolling deployments.

What to Analyze

1. Resource Specifications: Explicit requests (for scheduling) and limits (to prevent node exhaustion) for CPU/memory.
2. Health Probes: Distinct livenessProbe (restarts hung container), readinessProbe (controls traffic routing), and startupProbe (slow boots).
3. Pod Disruption & Graceful Termination: terminationGracePeriodSeconds, preStop lifecycle sleep hook, and PodDisruptionBudget.
4. Security Context Hardening: readOnlyRootFilesystem: true, allowPrivilegeEscalation: false, runAsNonRoot: true, dropping all capabilities (drop: [ALL]).
5. Autoscaling (HPA): HorizontalPodAutoscaler targets configured with appropriate min/max replicas.

Output Format

• Production Readiness Score: Checklist of Passed / Failed best practices.
• Hardened Kubernetes Manifest YAML: Complete corrected manifest.
• Helm Values / Template Fixes: Specific templating corrections.

Avoid

• Leaving CPU/Memory limits undefined (leads to noisy neighbor node crashes).
• Pointing liveness probes to the same endpoint as readiness probes with zero delay.