Triage Anton cluster Kubernetes health from CNI/DNS to apps.
npx skills add https://github.com/wcygan/anton --skill anton-cluster-health
Please help me install this Agent Skill. Skill: anton-cluster-health Source: https://github.com/wcygan/anton/tree/main/.claude/skills/anton-cluster-health Command: npx skills add https://github.com/wcygan/anton --skill anton-cluster-health
Provides a safe, read-only triage of the Kubernetes stack from the network and control planes down to applications, helping operators quickly determine if the cluster is healthy and where to start investigating.
Describe the current health of the Anton cluster by running the layer-3 to layer-6 triage and report the first red layer and recommended next steps.
High-intent search queries and answers about installing and using this skill.
Kubernetes cluster health triage runs read-only checks from the CNI/DNS layer up through Flux, cert-manager, and applications to surface the first failing layer. It reports the initial red layer and recommends targeted next steps for investigation.
Kubernetes triage explicitly checks silent-killer scenarios like 1Password ESO store readiness, envoy gateway programming, and cloudflared tunnel status to prevent false positives from Flux health checks alone. It identifies the actual failing component.
Yes, the triage performs strictly read-only checks across layers 3 through 6, ensuring no mutations to the Kubernetes cluster state. It safely determines if the Anton cluster is healthy and pinpoints where to start investigating failures.
Diagnosing Flux, cert-manager, or ESO issues requires a stepwise triage that checks the CNI/DNS network layer first, then platform components, and finally applications. This approach surfaces the first failing layer and points to the appropriate next debugging skill.
Applications can fail despite healthy Flux reconciliation due to underlying CNI/DNS issues or silent-killer scenarios like unready 1Password ESO stores or misconfigured envoy gateways. The triage checks these dependencies first before evaluating Flux status.
Read-only triage limitations include its inability to mutate cluster state or automatically resolve detected failures. It only identifies the first failing layer and recommends next steps, requiring a separate debugging skill to execute actual remediation.
dapperdiversDiagnose Kubernetes metrics-server failures and generate structured health reports.
wcyganDiagnose Flux reconciliation issues in Anton through an ordered diagnostic path.
jrmatherlyProvide deployment and runtime context for Spacebot's Talos Kubernetes cluster.
foxj77Diagnose cert-manager TLS certificate issues across Kubernetes resources.
