What problem does it solve? Product teams often start building on unverified assumptions, invented personas, and vague problem statements. This Skill runs a structured discovery and define session that gathers real material first, tags every claim with its evidence tier, and writes a product brief that downstream planning, spec-writing, and review skills can rely on. ## Core Features & Use Cases - Evidence-first brief generation: Writes product-brief.md with one claim per line, each tagged by evidence tier (interview, data, document, product, benchmark, synthetic, assumption) plus a coverage line. - Three discovery modes: Adapts question ladders, mandatory sections, and sign-off rules for existing products, client ideas, or the team's own ideas. - Quality and skeptic gates: Scores the draft against a quality gate and dispatches a fresh-context skeptic subagent whose findings return as plain interview questions. - Use Case: A founder with a new product idea runs the skill, answers two rounds of concrete questions, and receives a product brief with riskiest assumptions, kill criteria, and a collection plan for missing evidence. ## Quick Start Ask the agent to run ot-discover on your product idea, for example: run a discovery session for our new billing feature and write the product brief.
