Your monitoring tool shows numbers. Mine builds memory.
Most system monitors are obsessed with now. "CPU 82%." "RAM 6.2 GB." Cool. But your PC is not a screenshot. It's a story.
So I built a small engine inside PC Workman that does something different. Instead of just reading values, it collects them every second, quietly building a timeline of your machine. Then every 60 seconds something interesting happens: it compresses those 60 raw snapshots into one clean, meaningful average. Not just data. Context.
A mini pipeline, one file
This is handled by core/scheduler.py. Under the hood it behaves like a small pipeline:
- per-second sampling (live state)
- minute-level aggregation (trend)
- hooks for anomaly detection, process-behaviour tracking, and a future event system
A CPU jump for 1 second? Noise. A sustained pattern over minutes? Signal.
Most tools show you noise. I'm trying to surface signal. This is one of those invisible systems you never notice… until everything else starts making more sense. Still early, still rough — but this is where PC Workman starts feeling less like a tool and more like something that actually understands your machine.
What matters more in monitoring — raw numbers, or patterns over time?