Diagnose unhealthy Kubernetes workloads by analyzing cluster context and logs.
npx skills add https://github.com/blouargant/yoke-registry --skill k8s-triage
Please help me install this Agent Skill. Skill: k8s-triage Source: https://github.com/blouargant/yoke-registry/tree/main/skills/DevOps/k8s-triage Command: npx skills add https://github.com/blouargant/yoke-registry --skill k8s-triage
This Skill helps you quickly identify why a Kubernetes workload is unhealthy by organizing evidence from deployments, pods, events, and logs into a clear failure category.
Use the k8s-triage skill when you see kubectl or Kubernetes symptoms by asking for a triage plan for the affected namespace and workload, e.g., "Diagnose why my Kubernetes deployment is unhealthy in namespace staging and recommend the single next action."
High-intent search queries and answers about installing and using this skill.
To diagnose Kubernetes pods stuck in CrashLoopBackOff, gather and analyze cluster context, unhealthy pod details, recent events, and prior logs to classify the failure into categories like image/pull, probe, or application crash, then apply a safe next action.
Triage failing Kubernetes workloads across namespaces by executing read-only kubectl investigation to capture workload state and events, classifying the failure into image, scheduling, probe, network, permission, or application categories to pinpoint the root cause.
Find why a Kubernetes deployment is not ready by collecting evidence from deployments, pods, events, and logs to classify the issue into image/pull, scheduling, probe, configuration, network, permission, or application crash behavior.
You can run read-only kubectl triage safely in production, but the workflow enforces guardrails that prevent destructive actions like delete without explicit confirmation and avoids production namespace modifications without override.
After kubectl triage identifies an application crash or configuration error, it routes you toward a deeper debugging skill for actionable next steps and proposes exactly one safe next action with dry-run preference when possible.
eveldQuery Kubernetes resources to determine cluster state and resource health.
damacusInvestigate Kubernetes deployment failures and generate structured remediation reports.
nist0Manage Kubernetes clusters with kubectl commands for workloads, logs, and resources.
Leap0920Diagnose OMC/Claude-Code session failures by analyzing logs, traces, and state.
