problem-definition

To turn opportunity discovery insights into clear product problem statements, you extract structured inputs from discover-*.md artifacts or oral intake, then run a divergent exploration and multi-agent inquiry to produce a versioned define markdown document with measurable success criteria.

The best way to set measurable success criteria for a new product feature is to run a three-agent inquiry session—using advocate, challenger, and observer roles—on your discovery artifacts, which surfaces consensus and risks to inform automated RICE scoring and structured scope decisions.

You calculate RICE scores and capture scope decisions from discovery artifacts by ingesting discover-*.md files into a Define workflow, running divergent exploration across five angles, and generating a versioned define-v{ver}-v{N}-{date}.md output that preserves evidence levels and decision records.

Yes, you can define product problems using oral user inputs instead of discover markdown files. The workflow supports guided oral intake to populate the Define phase, running the same multi-agent inquiry and divergence drafts to produce structured problem statements and automated RICE scoring.

Multi-agent inquiry for product opportunity assessment works by deploying three distinct roles—advocate, challenger, and observer—to analyze discovery inputs. This mechanism surfaces consensus, identifies risks, and pulls industry references to ensure evidence-level transparency before finalizing problem definitions.

Product teams experience scope drift during the feature definition phase due to missing evidence-level transparency and unrecorded decisions. Enforcing iteration tracking, structured decision capture, and versioned define-*.md outputs prevents guesswork and aligns teams on what to build.