Diagnose Kubernetes and GitOps failures and propose production-safe infrastructure adjustments.
npx skills add https://github.com/nsvk13/skills --skill devops-core
Please help me install this Agent Skill. Skill: devops-core Source: https://github.com/nsvk13/skills/tree/main/devops-core Command: npx skills add https://github.com/nsvk13/skills --skill devops-core
Способность быстро и безопасно проектировать, развертывать и отлаживать инфраструктуру: от Kubernetes-кластеров и GitOps-потоков до CI/CD пайплайнов и IaC-подходов, когда нужно понять «что сломалось» и привести систему к рабочему продакшен-состоянию.
Пример: пользователь вставляет вывод kubectl/logs и описывает падение пода; Skill помогает сопоставить симптом с вероятной причиной (лимиты/ресурсы, сеть/политики, теги образов, несоответствие манифестов), после чего предлагает конкретный набор команд и корректировки манифеста/конвейера.
Use the devops-core skill to debug this Kubernetes issue based on the pasted kubectl output and logs, then propose the minimal manifest and CI/GitOps changes to fix it safely in production.
High-intent search queries and answers about installing and using this skill.
Troubleshoot Kubernetes CrashLoopBackOff and OOMKilled pods by reading events, logs, and resource limits to identify resource exhaustion or manifest mismatches. You get actionable diagnostic commands and manifest adjustments to safely resolve pod failures in production.
Set up production-ready Kubernetes manifests with FluxCD GitOps by configuring Deployment probes, RBAC, and securityContext defaults alongside HelmRelease and Kustomization. You get guarded GitOps reconciliation workflows for managed cluster updates.
Design GitLab CI/CD pipelines for partial rebuilds by structuring DAG-based job dependencies to skip unchanged stages. You get modular pipeline configurations that optimize execution time across different environments.
Yes, this approach supports modular Terraform and Ansible infrastructure provisioning by applying environment-specific IaC workflows with built-in guardrails. You get structured module layouts that integrate monitoring and safety constraints.
Kubernetes pods stuck in Pending or ImagePullBackOff states often result from insufficient cluster resources, network policies, or incorrect image tags. You get targeted diagnostic steps to verify resource scheduling and correct manifest deployments.
Yes, you can use PromQL to define SLOs and alerts for Prometheus monitoring by writing queries that track application latency and error rates. You get actionable PromQL adjustments and Grafana dashboard configurations for production observability.
openeuler-mirrorCorrelate kernel logs, process memory snapshots, and cgroup metrics to diagnose Linux OOM events.
ringier-dataDeploy package updates to FluxCD environments via gitops manifests and reconciliation.
alexei-ledCodify Kubernetes, Terraform, Helm, Kustomize, and GitHub Actions configuration patterns.
shafibabarDefine GitOps operational models with Git as the source of truth for Kubernetes.
