Claude Skill

kubernetes

Operate Kubernetes with kubectl, Helm, k9s, kustomize, and stern.

LLM Mart · 0 points · 4 views 0 listing impressions 0 install-command copies
Download fmind-dot-skills_kubernetes-3bbe830.zip · 2 KB

Install

skills CLI npx skills add https://github.com/fmind/dot/tree/main/skills/kubernetes
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install fmind-dot@llmmart
Git git clone https://github.com/fmind/dot.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole fmind/dot collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Kubernetes Cluster and Workload Operations

Inspect clusters, validate manifests, deploy Helm/Kustomize releases, and read bounded logs with kubectl, helm, kustomize, kubeconform, and stern (docker owns container runtimes; infra-as-code owns cloud cluster provisioning).

None of these tools (or k3d) are installed globally: pin required ones in the project's mise.toml via mise use --pin <tool> (mise). Confirm authority before mutating clusters or incurring spend, and pass --context and --namespace (--kube-context in Helm) explicitly on every call.

Workflow

  1. Confirm context and cluster: never assume default kubeconfig points to dev:

    kubectl config get-contexts
    kubectl --context <context> cluster-info
    
  2. Validate manifests offline: check target-version schemas (kubeconform defaults to master) and scan with trivy before applying (prefer offline validation over spinning up k3d when sufficient):

    kubeconform -strict -summary -kubernetes-version <cluster-version> <file-or-directory>
    trivy --config trivy.yaml config <file-or-directory>
    kustomize build <kustomization-dir> | kubeconform -strict -summary -kubernetes-version <cluster-version> -
    
  3. Preview and deploy declaratively:

    helm upgrade --install <release-name> <chart-path> --kube-context <context> --namespace <namespace> --dry-run=server --hide-secret
    kubectl --context <context> --namespace <namespace> diff -k <kustomization-dir>
    

    kubectl diff exits 1 on diffs and >1 on errors; it masks Secret data (without --show-secrets) but not ConfigMaps or rendered values. On authorized Helm 4 applies, pass --rollback-on-failure --timeout <duration> (replaces deprecated --atomic; --force-replace replaces --force).

  4. Read bounded logs and events (--tail applies per container; reserve streaming stern for interactive sessions):

    stern <pod-query> --context <context> -n <namespace> --since 15m --tail 50 --no-follow
    kubectl --context <context> get pods,events -n <namespace>
    

Gotchas

  • Never mutate persistent kubeconfig context unless explicitly asked; pass --context per command.
  • Never dump raw Secret values: inspect keys/sizes via kubectl describe secret <name> instead of kubectl get secret -o yaml.

Task guides

  • k3d: Create, budget, and tear down disposable local k3d clusters for runtime tests.

Documentation

Files (dot)
  • references
    • k3d.md 1 KB
      Loading…
  • SKILL.md 3.3 KB
    ---
    name: kubernetes
    description: "Operate Kubernetes with kubectl, Helm, kustomize, kubeconform, and stern."
    license: MIT
    metadata:
      kind: task
      author: Médéric HURIER (Fmind)
      source: github.com/fmind/dot/tree/main/skills/kubernetes
      created: "2026-09-16"
      updated: "2026-11-10"
    ---
    
    # Kubernetes Cluster and Workload Operations
    
    Inspect clusters, validate manifests, deploy Helm/Kustomize releases, and read bounded logs with `kubectl`, `helm`, `kustomize`, `kubeconform`, and `stern` ([docker](../docker/SKILL.md) owns container runtimes; [infra-as-code](../infra-as-code/SKILL.md) owns cloud cluster provisioning).
    
    None of these tools (or `k3d`) are installed globally: pin required ones in the project's `mise.toml` via `mise use --pin <tool>` ([mise](../mise/SKILL.md)). Confirm authority before mutating clusters or incurring spend, and pass `--context` and `--namespace` (`--kube-context` in Helm) explicitly on every call.
    
    ## Workflow
    
    1. **Confirm context and cluster**: never assume default kubeconfig points to dev:
    
       ```bash
       kubectl config get-contexts
       kubectl --context <context> cluster-info
       ```
    
    1. **Validate manifests offline**: check target-version schemas (`kubeconform` defaults to `master`) and scan with [trivy](../code-security/references/trivy/GUIDE.md) before applying (prefer offline validation over spinning up [k3d](references/k3d.md) when sufficient):
    
       ```bash
       kubeconform -strict -summary -kubernetes-version <cluster-version> <file-or-directory>
       trivy --config trivy.yaml config <file-or-directory>
       kustomize build <kustomization-dir> | kubeconform -strict -summary -kubernetes-version <cluster-version> -
       ```
    
    1. **Preview and deploy declaratively**:
    
       ```bash
       helm upgrade --install <release-name> <chart-path> --kube-context <context> --namespace <namespace> --dry-run=server --hide-secret
       kubectl --context <context> --namespace <namespace> diff -k <kustomization-dir>
       ```
    
       `kubectl diff` exits `1` on diffs and `>1` on errors; it masks `Secret` data (without `--show-secrets`) but not `ConfigMap`s or rendered values. On authorized Helm 4 applies, pass `--rollback-on-failure --timeout <duration>` (replaces deprecated `--atomic`; `--force-replace` replaces `--force`).
    
    1. **Read bounded logs and events** (`--tail` applies per container; reserve streaming `stern` for interactive sessions):
    
       ```bash
       stern <pod-query> --context <context> -n <namespace> --since 15m --tail 50 --no-follow
       kubectl --context <context> get pods,events -n <namespace>
       ```
    
    ## Gotchas
    
    - **Never mutate persistent kubeconfig context** unless explicitly asked; pass `--context` per command.
    - **Never dump raw Secret values**: inspect keys/sizes via `kubectl describe secret <name>` instead of `kubectl get secret -o yaml`.
    
    ## Task guides
    
    <!-- guides:start -->
    
    - [k3d](references/k3d.md): Create, budget, and tear down disposable local k3d clusters for runtime tests.
    
    <!-- guides:end -->
    
    ## Documentation
    
    - [Kubernetes](https://kubernetes.io/docs/) · [Helm](https://helm.sh/docs/) · [Stern](https://github.com/stern/stern) · [kubeconform](https://github.com/yannh/kubeconform) · [Kubernetes releases](https://github.com/kubernetes/kubernetes/releases) · [Helm releases](https://github.com/helm/helm/releases)
    - Companion skills: [docker](../docker/SKILL.md), [infra-as-code](../infra-as-code/SKILL.md).
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.