Use when an epic or release is split into testable areas and someone must commit test capacity. Produces a PERT three-point estimate as ranges, with a named assumptions ledger and a split across developer, automation and exploratory work; also classifies a change set's shape from git history.