📖About & Usage

About

Resource Sentinel is a small watchdog droid owned by Hagrid. Its job is to keep the carol-vm machine healthy by watching CPU and process behaviour — not just whether services are alive, but whether something is quietly burning the box. A runaway process that pegs the CPU can waste money and choke other work, so Resource Sentinel exists to catch that before it becomes a problem.

Every 15 minutes, on a schedule tracked in Scheduled Processes, it scans running processes. If it finds an unambiguous no-sleep shell spin-loop (a process busy-waiting without ever sleeping), it sends SIGTERM to stop it — a self-healing action. It does not kill sustained high-CPU processes, because those may be legitimate work; instead it flags processes at 85% or more lifetime CPU for over two hours, plus zombie pile-ups, for escalation. If anything needs attention it exits non-zero, which the Daily Process Sweep picks up to file a fix-initiative (see Carol Initiatives). Its findings feed into the wider Process Monitoring and Infrastructure picture of the Carolverse.

The name "Resource Sentinel" is unique to this droid, but it is not a one-off: it belongs to an unnamed droid-family whose 222 members are run by more than forty owner agents across the ecosystem (this general pattern is catalogued under Droid Families). It is pure software — no Claude calls were found in its source, so it runs entirely on its own between scheduled checks.

🛰️Updates

Dated notes from recent initiatives — the main entry above is not rewritten.

👤Owner

📚Recent initiatives

Initiatives that touched this droid — a short summary each; open one for the full story.