What problem does it solve? Product bets often rest on unchecked beliefs, and teams default to building an MVP to find out if they are right — an expensive, slow way to learn. This Skill gives agents a disciplined method to identify the riskiest assumption behind a bet and resolve it with a cheap, fast test before any build work begins. ## Core Features & Use Cases - Four Assumption Categories: Classify beliefs as desirability, viability, feasibility, or usability, then rank them by risk and confidence to pick what to test first. - Pre-Registration Discipline: A structured template forces a disconfirming signal and concrete pass/fail criteria to be written before the test runs, preventing confirmation-seeking. - Fast-Test Shapes: Choose from data pulls, 5-user interviews, concept tests, smoke tests, wizard-of-oz tests, and engineering spikes — all designed to resolve in days. - Use Case: A stakeholder proposes building an MVP to see if SMB users want self-serve data export. Instead, run a 4-day moderated concept test with 5 admins and pre-registered thresholds to decide whether to proceed, reframe, or kill the bet. ## Quick Start Ask the agent to design an assumption test for the riskiest belief behind your current product bet, with pre-registered pass/fail criteria that resolve within days.
