discovery

The product discovery process works by conducting deep conversational interviews with no round limit, asking targeted follow-ups until every behavioral decision belongs to the operator, and then generating a sub-one-page brief saved to .pi/specs/<feature>/brief.md.

To run user interviews for product discovery without assumptions, the Skill probes open-ended operator descriptions iteratively, asking targeted follow-up questions until no assumptions remain and all feature requirements are explicitly defined.

You need a Gate 1 discovery process for feature requirements when a product or feature requires greenfield or brownfield work, ensuring you capture a single-source, assumption-free brief before proceeding to the next phase.

Yes, you can generate a kebab-case feature ID and update workflow-state.json automatically, as the Skill handles structured output by saving the brief to .pi/specs/<feature>/brief.md and updating workflow-state.json and .pi/feature-list.json accordingly.

The best way to conduct deep conversational interviews for product requirements is using an iterative approach that handles open-ended descriptions, probes until no assumptions remain, and optionally incorporates reference research to produce a formal brief document.

Yes, the Gate 1 brief presentation supports approval, correction, or restart decisions by presenting the brief in Portuguese and collecting a formal gate confirmation to drive the next workflow phase.