User and Expert Interview Research Assistant

Problem It Solves

User interviews often become fragmented: outlines, recruitment, screening, notes, and reports are handled separately, so user quotes are lost and only abstract conclusions remain. For engineers and PMs, the challenge is not merely formatting text, but turning behavior, workarounds, and conflicting feedback into traceable findings.

How It Works

The skill routes input signals into five phases: Phase 1 interview guide, Phase 2 recruitment requirements, Phase 3 candidate screening, Phase 4 interview notes, and Phase 5 research report. It reads the corresponding references/*.md files before executing, rather than relying on memory. Core principles include user quotes first, behavioral facts first, workarounds as demand signals, and traceable recommendations. It also maintains shared context such as research goals, user personas, core questions, and screening criteria, and can start from any phase; ambiguous inputs are confirmed before proceeding.

Boundaries

It is useful for interview projects with transcripts, resumes, research goals, or recruitment criteria, and produces stakeholder-ready output. If source records or key context are missing, inference may be incomplete; preferences such as anonymization and detail level should be confirmed at the start. It does not replace conducting interviews, managing incentives, or compliance review.