host_resource_diagnosis

To diagnose CPU spikes and memory pressure on a host machine, you can collect real-time metrics using psutil to analyze top processes by resource usage and identify root causes. This generates an evidence-backed report without modifying system state.

The best way to identify suspected OOM conditions in a container is to run a staged diagnostic flow that analyzes memory metrics and top processes. This approach isolates the memory pressure root cause safely without altering the system state.

Yes, psutil works for diagnosing disk fullness on local on-host systems by collecting real-time disk metrics. The skill analyzes these metrics alongside CPU and memory data to safely identify resource bottlenecks and output intervention suggestions.

To generate an evidence-backed resource diagnosis report without changing system state, collect local CPU, memory, and disk metrics via psutil. The diagnostic flow analyzes top processes and outputs a safe, reproducible report for audit and intervention.

Yes, you can use this resource diagnosis approach for both on-host machines and containers. It collects local psutil data to analyze CPU, memory, and disk metrics, effectively identifying resource pressure and OOM conditions across both environments.

You should not use a read-only resource diagnosis when you need to actively resolve issues by killing processes or modifying system state. This approach strictly collects metrics and generates reports, providing intervention suggestions without executing changes.