Changelog

0.1.0

First release. The oee-ops skill: the NF E 60-182 time hierarchy and the three rates, nine documented ways an OEE comes out wrong while looking reasonable, interval-set arithmetic for the time buckets, loss waterfall and Pareto by cause, a shift analyser CLI, and a deterministic benchmark on round numbers.

impactOfRemoving is labelled a model rather than a measurement: removing a stop alone moves TRS by nothing, so the projected gain rests on the assumption that recovered time runs at the window's own performance and quality. Any tool that reports a gain is making that assumption whether or not it says so.