Skill
Community

infrastructure-management

Install
1
Install the plugin
$
npx claudepluginhub levifig/loaf --plugin loaf

Want just this skill?

Then install: npx claudepluginhub u/[userId]/[slug]

Description

Covers Docker, Kubernetes, GitOps, CI/CD pipelines, and container security.

Tool Access

This skill is limited to using the following tools:

ReadWriteEditBashGlobGrep
Supporting Assets
View in Repository
references/ci-cd.md
references/docker.md
references/gitops.md
references/kubernetes.md
references/troubleshooting.md
scripts/check-dockerfile.sh
scripts/validate-k8s-manifest.py
Skill Content

Infrastructure

Infrastructure patterns for containerization, orchestration, CI/CD pipelines, and deployment automation.

Stack Overview

LayerTechnologies
ContainersDocker, BuildKit, multi-stage builds
OrchestrationKubernetes, Helm, Kustomize
GitOpsArgoCD, Flux, Argo Rollouts
CI/CDGitHub Actions, GitLab CI
RegistriesGHCR, ECR, GCR, DockerHub

Topics

TopicReference FileUse When
Dockerreferences/docker.mdWriting Dockerfiles, optimizing builds, adding health checks
Kubernetesreferences/kubernetes.mdCreating deployments, services, probes, resource limits
GitOpsreferences/gitops.mdSetting up ArgoCD, Kustomize, sync policies
CI/CDreferences/ci-cd.mdBuilding GitHub Actions workflows, caching, secrets
Troubleshootingreferences/troubleshooting.mdDebugging CI failures, version conflicts, cache issues

Available Scripts

ScriptUsageDescription
scripts/check-dockerfile.shcheck-dockerfile.sh <file>Validate Dockerfile best practices
scripts/validate-k8s-manifest.pyvalidate-k8s-manifest.py <file>Check K8s manifest for required fields

Critical Rules

Always

  • Use multi-stage builds to minimize image size
  • Run containers as non-root user (UID 1000)
  • Include health checks in all services
  • Pin specific image versions (no :latest)
  • Set resource requests AND limits
  • Use npm ci / pip-sync in CI (not install)
  • Commit lockfiles to version control

Never

  • Commit secrets to version control
  • Use :latest tags in production
  • Skip security scanning in CI
  • Deploy without rollback capability
  • Store state in containers
  • Run as root in production

CI Failure Triage

CI Failed
+-- Same code passes locally?
|   +-- YES --> Check environment differences
|   |   +-- Python/Node version
|   |   +-- Environment variables
|   |   +-- File permissions
|   |   +-- Installed dependencies
|   +-- NO --> Fix the actual bug
+-- Flaky (sometimes passes)?
|   +-- Check for race conditions, shared state, timeouts
+-- Always fails in CI?
    +-- Check runner resources (memory, timeout)
    +-- Check external service access
    +-- Check CI-specific config
Stats
Stars1
Forks0
Last CommitFeb 21, 2026

Similar Skills