What problem does it solve? Operating CMDB environments (dev/test/stage/loadtest via docker-compose and production Kubernetes clusters via kube_ci) requires strict, error-prone operational procedures where improvisation can cause production data loss. This Skill acts as a pure runbook executor that deploys, backs up, restores, and troubleshoots environments without ever modifying code, configs, or scripts. ## Core Features & Use Cases - Environment lifecycle management: Setup, deploy, backup, restore, and troubleshoot docker-compose environments (dev/test/stage/loadtest) through the cmdb-env.sh script with mandatory healthchecks. - Kubernetes production operations: Deploy, redeploy, dismiss, backup, and restore CMDB instances (prod-adp2, prod-testop, preprod targets) exclusively through kube_ci pipeline scripts on the KVM host, with enforced backup and architect confirmation before destructive production actions. - Production safety guardrails: Hard NEVER rules prevent direct kubectl apply/helm install, namespace deletion without backup, scp of repo files, and any file modification, with explicit escalation paths to other roles. - Use Case: Run a full production redeploy of the prod-adp2 CMDB instance: the Skill resolves the target, takes a pg_dump backup, asks for architect confirmation, deletes the namespace, redeploys via kube_ci, and verifies health and login. ## Quick Start Ask the agent to deploy the stage environment or run a Kubernetes backup of prod-adp2, for example by invoking the ops-engineer skill with the argument "deploy stage" or "k8s-backup prod-adp2".
