Help us improve
Share bugs, ideas, or general feedback.
From cce-kubernetes
Runs Kubernetes cluster health diagnostics using dynamic API discovery. Use for checking cluster health, troubleshooting K8s issues, or health assessments.
npx claudepluginhub nodnarbnitram/claude-code-extensions --plugin cce-kubernetesHow this skill is triggered — by the user, by Claude, or both
Slash command
/cce-kubernetes:kubernetes-healthThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
> Dynamic, discovery-driven health checks for any Kubernetes cluster configuration
Diagnoses Kubernetes cluster health issues (pods, deployments, resources, events) and performs user-approved security fixes. Useful for incident response, on-call debugging, and cluster troubleshooting.
Diagnoses and fixes Kubernetes pod failures like CrashLoopBackOff, Pending, DNS, networking, storage mounts, and rollout issues using kubectl workflows and scripts.
Manages Kubernetes operations: deployments, workloads, networking, storage, troubleshooting, kubectl mastery, and cluster stability for pods and services.
Share bugs, ideas, or general feedback.
Dynamic, discovery-driven health checks for any Kubernetes cluster configuration
| Impact | Value |
|---|---|
| Token Savings | ~70% vs manual kubectl exploration |
| Setup Time | 0 min (uses existing kubectl config) |
| Coverage | Adapts to installed operators automatically |
| Problem | Root Cause | How This Skill Helps |
|---|---|---|
| Missing operator health | Static checklists miss CRDs | Dynamic API discovery detects all installed operators |
| Stale diagnostics | Manual checks become outdated | Real-time cluster API interrogation |
| Incomplete coverage | Unknown cluster configuration | Automatically activates relevant sub-agents |
kubectl is configured and can reach your clusterdiscover_apis.py to detect installed operators# Step 1: Verify kubectl context
kubectl config current-context
kubectl cluster-info
# Step 2: Run API discovery
uv run ${CLAUDE_PLUGIN_ROOT}/skills/kubernetes-health/scripts/discover_apis.py
# Step 3: Review detected operators and dispatch health agents
| Mistake | Why It's Wrong | Correct Approach |
|---|---|---|
| Hardcoding operator checks | Misses installed operators, checks missing ones | Use API discovery to detect what's installed |
| Sequential agent dispatch | Slow for multi-operator clusters | Run operator agents in parallel (same priority) |
| Raw kubectl output | Token inefficient, hard to parse | Use scripts for condensed JSON output |
| Script | Purpose |
|---|---|
scripts/discover_apis.py | Discovers all API groups and detects installed operators |
scripts/health_orchestrator.py | Maps discovered APIs to specialized health agents |
scripts/aggregate_report.py | Aggregates multi-agent results into unified report |
| File | Contents |
|---|---|
references/operator-checks.md | Detailed health checks for each supported operator |
references/health-scoring.md | Scoring methodology and weight assignments |
| File | Purpose |
|---|---|
templates/health-report.json | JSON schema for health report output |
| Package | Version | Purpose |
|---|---|---|
| kubectl | Latest | Cluster interaction |
| Python | >= 3.11 | Script execution |
| uv | Latest | Python script runner |
| Package | Version | Purpose |
|---|---|---|
| kubernetes | >= 28.1.0 | Python client (for advanced discovery) |
The skill automatically detects and dispatches specialized agents for:
| Operator | API Group | Agent |
|---|---|---|
| Core K8s | (always) | k8s-core-health-agent |
| Crossplane | crossplane.io | k8s-crossplane-health-agent |
| ArgoCD | argoproj.io | k8s-argocd-health-agent |
| Cert-Manager | cert-manager.io | k8s-certmanager-health-agent |
| Prometheus | monitoring.coreos.com | k8s-prometheus-health-agent |
| Status | Score Range | Criteria |
|---|---|---|
| HEALTHY | 90-100 | All checks pass, no warnings |
| DEGRADED | 60-89 | Some warnings, no critical issues |
| CRITICAL | 0-59 | Critical issues affecting availability |
# Verify context
kubectl config current-context
# Test connectivity
kubectl cluster-info
# Check permissions
kubectl auth can-i get pods --all-namespaces
${CLAUDE_PLUGIN_ROOT}/agents/specialized/kubernetes/${CLAUDE_PLUGIN_ROOT}/agents/specialized/kubernetes/