一键导入
gke-pod-restarts
Use this skill to investigate GKE incidents involving pod restart cascades, 503/504 errors, service unavailability, or CrashLoopBackOff. Triggers: 'pods keep restarting', '503/504 errors', 'service is down', 'crash loop', 'what caused the incident at <time>'. Inputs required: GCP project, cluster name+region, namespace, service/deployment names, approximate incident time + timezone. All commands are read-only. Never mutate cluster state.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。