system-monitoring

name	system_monitoring
description	Monitor system resource usage (CPU, RAM, GPU) and active processes on CachyOS.

System Monitoring Skill

This skill teaches the Hermes Agent how to monitor active system hardware loads on CachyOS.

Triggering Phrases

• "Show my system resource usage"
• "Is my GPU currently being utilized by Ollama?"
• "List the top CPU-consuming processes"

Instructions

When the user requests resource utilization data or asks about system loads, run the following commands to gather metrics:

1. Check Memory (RAM & Swap) usage:

   free -h
   

2. Check System Load & CPU utilization:

   uptime
   # Or get the top CPU-consuming processes
   ps -eo pid,cmd,%mem,%cpu --sort=-%cpu | head -n 10
   

3. Check GPU utilization:

   • For NVIDIA GPUs:

     nvidia-smi --query-gpu=utilization.gpu,utilization.memory,temperature.gpu,memory.used,memory.total --format=csv
     

   • For AMD GPUs:

     cat /sys/class/drm/card0/device/gpu_busy_percent 2>/dev/null || echo "ROCm GPU telemetry not available directly via sysfs."
     

4. Verify if Ollama is running active processes:

   pgrep -fl ollama
   

5. Provide a summarized report to the user: Report the active memory pool details, CPU load factors, and GPU telemetry details back in a clear, formatted summary.