enterprise-agent-ops

Operational controls for long-lived or cloud-hosted agent systems — runtime lifecycle (start, pause, stop, restart), observability (logs, metrics, traces), least-privilege safety scopes and kill switches, and rollout/rollback change management with audit logs and success/cost metrics. Use when running production agent fleets on PM2, systemd, or containers that need monitoring, incident response, or deployment gates.