Diagnose unhealthy Kubernetes resources and GitOps workflows with read-only commands.
npx skills add https://github.com/confighub/cub-scout --skill scout-diagnose
Please help me install this Agent Skill. Skill: scout-diagnose Source: https://github.com/confighub/cub-scout/tree/main/skills/scout-diagnose Command: npx skills add https://github.com/confighub/cub-scout --skill scout-diagnose
cub-scout helps you interpret why Kubernetes or GitOps state looks unhealthy and decide what to do next, without making any changes.
Ask: “Explain the Deployment api in namespace prod and tell me what’s wrong and the next read-only command I should run to investigate.”
High-intent search queries and answers about installing and using this skill.
Diagnose unhealthy Kubernetes resources by applying read-only kubectl commands to observe live cluster state and match symptoms against known failure patterns. This generates structured diagnosis outputs with suggested next-step remediation commands while strictly avoiding any mutations to cluster state.
GitOps reconciliation gets stuck due to source synchronization errors or application health failures in pipelines like Flux or Argo. Diagnose these GitOps issues by analyzing pipeline health status and matching stuck reconcile symptoms against built-in GitOps failure patterns to identify the root cause.
Yes, you can debug Kubernetes cluster issues from offline scan outputs or file inputs. The diagnosis process accepts live cluster observations as well as offline file inputs to interpret symptoms, explain errors, and recommend next-step investigation commands without requiring direct cluster connectivity.
Pattern matching for Kubernetes failure detection is the process of comparing observed cluster symptoms against a library of known issue patterns. It translates observed error states into actionable diagnosis guidance by identifying which built-in Kubernetes or GitOps failure mode matches your current unhealthy resource state.
No, read-only remediation for GitOps workflows does not allow applying patches or any mutations to cluster state. It restricts actions strictly to kubectl and GitOps read surfaces, focusing entirely on generating structured outputs with suggested remedies and next-step investigation commands rather than executing changes.
vesvietDebug Kubernetes deployment issues including pod failures and ArgoCD sync problems.
dapperdiversDiagnose Kubernetes metrics-server failures and generate structured health reports.
keiretsu-labsDiagnose and manage Flux CD installations across multiple Kubernetes clusters.
jctannerCompare GENERATED_ARCHITECTURE.md against live Kubernetes cluster resources and report drift.
