What problem does it solve? Teams write epics as plans and commitments, so weak ideas get built because nobody ever wrote down the result that would make them stop. This Skill turns an epic into a bet the team can actually lose: a target user, a behaviour change with a baseline, a confirming signal, a falsifying signal, and the cheapest test that could produce either. ## Core Features & Use Cases - Falsifiable hypothesis framing: Renders the epic in both If/Then and Gothelf Lean UX grammars, then forces a numeric falsifying threshold with a named owner who can act on it. - Signal design with guardrails: Builds a signal table (Measure, Baseline, Source, Confirm, Falsify, Window) that rejects adoption metrics, derives the window from the user journey rather than the sprint, and requires at least one guardrail metric. - Cheapest-test selection: Climbs a validation ladder from existing data queries up to an instrumented thin slice, ruling out cheaper rungs explicitly. - Use Case: Before splitting or prioritising an epic, a product manager runs this Skill to produce an Epic Hypothesis card to paste into Jira, ADO, or Linear, including a pre-committed decision rule and a read-out date. ## Quick Start Frame this epic as a falsifiable hypothesis with confirming and falsifying signals, a decision rule, and the cheapest test that could prove it wrong.
