What problem does it solve? Discovery evidence arrives fragmented across interview notes, survey comments, support themes, transcripts, and analytics summaries, making it hard to produce a coherent input for discovery and PRD work without losing contradictions or overstating certainty. ## Core Features & Use Cases - Evidence Synthesis: Normalize mixed sources, surface converging themes, contradictions, and gaps, and label evidence strength qualitatively (Strong, Moderate, Directional, Unknown). - Job Verdicts & New-Job Signals: Render a supports/challenges/suggests-refinement verdict per active JTBD with cited evidence, and record raw new-job signals routed to the jobs-to-be-done skill. - Durable Artifacts: Write append-only synthesis files to docs/product/research/<date>-<slug>.md with a sources frontmatter, plus a context.md digest and [conflict] ledger entries. - Use Case: After six admin interviews, a support-tag summary, and funnel analytics, produce one synthesis showing permission hesitation as the top theme, a refinement verdict on the onboarding job, and discovery questions to carry forward. ## Quick Start Synthesize the un-ingested batches in docs/product/input/ into a research synthesis answering why first-time admins abandon workspace setup.
