kubernetes-deployment-automation

license	Apache-2.0
name	kubernetes-deployment-automation
description	Automate Kubernetes deployment lifecycle: kubectl rollouts, GitOps sync (ArgoCD/Flux), namespace and secret management, rollout monitoring, and failed deployment troubleshooting. Use when deploying manifests to clusters, debugging stuck rollouts, setting up GitOps pipelines, or managing multi-environment promotion. NOT for: generating K8s manifests (use kubernetes-manifest-generator), blue-green/canary traffic shifting (use blue-green-deployment-orchestrator), cluster provisioning (use terraform-module-builder), CI/CD pipeline config (use github-actions-pipeline-builder).
allowed-tools	Read,Write,Edit,Bash(kubectl:*,helm:*,argocd:*,flux:*,kustomize:*)
metadata	{"category":"DevOps & Infrastructure","tags":["kubernetes","deployment","gitops","argocd","flux","rollout","operations"],"pairs-with":[{"skill":"kubernetes-manifest-generator","reason":"Generates the manifests this skill deploys"},{"skill":"blue-green-deployment-orchestrator","reason":"Handles progressive delivery strategies after initial deployment"},{"skill":"devops-automator","reason":"Broader CI/CD context that triggers deployments"},{"skill":"secret-management-expert","reason":"Secrets must be provisioned before workloads deploy"},{"skill":"monitoring-stack-deployer","reason":"Observability required to validate deployment health"}],"argument-hint":"[action: deploy|rollback|troubleshoot|gitops-setup|promote]"}

Kubernetes Deployment Automation

Operational expertise for the deployment lifecycle — from kubectl apply to production-verified rollout. Owns Day 2 operations: deploying manifests, monitoring rollouts, troubleshooting failures, managing secrets/namespaces, and GitOps workflows.

When to Use

Use for:

• Deploying manifests to Kubernetes clusters (kubectl, Helm, Kustomize)
• Setting up GitOps with ArgoCD or Flux
• Monitoring and verifying rollout health
• Troubleshooting stuck or failed deployments
• Multi-environment promotion (dev → staging → prod)
• Namespace, RBAC, and secret lifecycle for deployments
• Rollback decisions and execution

NOT for:

• Generating K8s YAML/Helm charts → kubernetes-manifest-generator
• Blue-green/canary traffic shifting → blue-green-deployment-orchestrator
• Cluster provisioning with Terraform → terraform-module-builder
• CI/CD pipeline YAML → github-actions-pipeline-builder
• Application-level debugging → language-specific skills

Core Process

flowchart TD
  START[Deployment Request] --> ASSESS{What stage?}
  ASSESS -->|New deployment| PREFLIGHT[Pre-flight Checks]
  ASSESS -->|Stuck rollout| TROUBLESHOOT[Troubleshoot]
  ASSESS -->|Setup GitOps| GITOPS[GitOps Setup]
  ASSESS -->|Promote env| PROMOTE[Promotion Pipeline]

  PREFLIGHT --> NS[Namespace exists?]
  NS -->|No| CREATE_NS[Create namespace + RBAC]
  NS -->|Yes| SECRETS[Secrets provisioned?]
  CREATE_NS --> SECRETS
  SECRETS -->|No| PROVISION_SECRETS[Provision secrets]
  SECRETS -->|Yes| DEPS[Dependencies ready?]
  PROVISION_SECRETS --> DEPS
  DEPS -->|No| WAIT_DEPS[Wait / deploy deps first]
  DEPS -->|Yes| DEPLOY[Deploy]

  DEPLOY --> METHOD{Deployment method?}
  METHOD -->|kubectl| KUBECTL[kubectl apply -f / rollout]
  METHOD -->|Helm| HELM[helm upgrade --install]
  METHOD -->|ArgoCD| ARGOCD[argocd app sync]
  METHOD -->|Flux| FLUX[flux reconcile]

  KUBECTL --> VERIFY[Verify Rollout]
  HELM --> VERIFY
  ARGOCD --> VERIFY
  FLUX --> VERIFY

  VERIFY --> HEALTHY{Healthy?}
  HEALTHY -->|Yes| DONE[Deployment Complete]
  HEALTHY -->|No| TROUBLESHOOT

  TROUBLESHOOT --> TRIAGE{Failure type?}
  TRIAGE -->|ImagePullBackOff| IMG[Check image tag, registry auth, pull policy]
  TRIAGE -->|CrashLoopBackOff| CRASH[Check logs, resource limits, probes]
  TRIAGE -->|Pending| PENDING[Check node resources, PVC, scheduling]
  TRIAGE -->|OOMKilled| OOM[Increase memory limits, check for leaks]
  TRIAGE -->|Timeout| TIMEOUT[Check readiness probe, startup time]
  IMG --> DECIDE_ROLLBACK{Fixable quickly?}
  CRASH --> DECIDE_ROLLBACK
  PENDING --> DECIDE_ROLLBACK
  OOM --> DECIDE_ROLLBACK
  TIMEOUT --> DECIDE_ROLLBACK
  DECIDE_ROLLBACK -->|Yes| FIX[Fix and redeploy]
  DECIDE_ROLLBACK -->|No| ROLLBACK[kubectl rollout undo]
  FIX --> DEPLOY
  ROLLBACK --> DONE