Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
A bash script that checks macOS system resources and flags processes exceeding memory, CPU, disk, or runtime thresholds. Outputs structured JSON to stdout, ready to pipe into other tools or feed to an agent. Common targets include stale AI coding agents, speech-to-text servers, and leaked dev processes.
Outputs machine-readable JSON instead of human-readable text, so results can be consumed directly by agents, scripts, or monitoring stacks without any parsing.
When to use it
- Catching a Claude or Codex agent still running after a long session
- Finding a Python script that leaked overnight and is consuming RAM
- Checking available disk space before starting a large Docker build
- Feeding structured process data into a custom alerting pipeline
- Scheduling a cron job to surface resource waste automatically
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Are there any wasteful or suspicious processes running on my machine?
- 1Run check.sh and capture stdout
- 2Parse the JSON output
- 3Check the suspicious field — if false, report the system is healthy
- 4If suspicious is true, extract issues and top_processes from the JSON
- 5Format a concise report listing each flagged process with PID, RAM usage, CPU, and uptime
A formatted report showing system resource summary and flagged processes, with suggested kill commands for stale or high-usage offenders