pre-build-interview

Validates feature ideas through structured pre-build user interviews ending in a build, iterate, or kill recommendation.

11|Updated Mar 20, 2026
One-click install
npx skills add https://github.com/qa-aman/next-leap-claude-code --skill pre-build-interview-qa-aman
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pre-build-interview
Source: https://github.com/qa-aman/next-leap-claude-code/tree/main/.claude/skills/pre-build-interview
Command: npx skills add https://github.com/qa-aman/next-leap-claude-code --skill pre-build-interview-qa-aman

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Teams routinely ship features users never adopt because interviews pitch the idea instead of testing real behavior, and there is no falsifiable hypothesis to fail against. This Skill runs a disciplined five-phase validation workflow that converts user interviews into an evidence-based build, iterate, or kill decision before engineering invests. ## Core Features & Use Cases - Hypothesis Definition: Forces a written one-page hypothesis with target segment, current workaround, signal thresholds, and kill criteria before any interviews begin. - Mom Test Interview Guides: Generates 8-12 question scripts that probe past behavior and current workarounds, automatically rewriting leading or hypothetical questions. - Decision Synthesis: Aggregates interviews against thresholds, maps JTBD Forces of Progress (push, pull, anxiety, habit), and builds a mini Opportunity Solution Tree to produce a BUILD, ITERATE, or KILL verdict. - Use Case: Before writing a PRD for a Smart Follow-Up feature, run 8-12 interviews with target users, capture their actual workarounds and switching anxieties, and receive a decision document stating whether to build, reshape, or kill the idea. ## Quick Start Ask the assistant to validate whether a planned feature is worth building by running pre-build user interviews before writing the PRD.

Frequently Asked Questions about pre-build-interview

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate a feature idea before building it?▼

Run pre-build user interviews with a written hypothesis, signal thresholds, and kill criteria before any interviews start. Interview 8-12 target users about past behavior and current workarounds, then score results against thresholds to reach a build, iterate, or kill decision.

What is the Mom Test for user interview questions?▼

The Mom Test is a question design method that asks about specific past behavior instead of hypotheticals or opinions. Questions like "would you use this" get rewritten into "tell me about the last time you tried to do X" to avoid polite false positives.

How many user interviews are enough for feature validation?▼

Eight to twelve interviews is the recommended range for problem validation. Below six you are reading noise rather than patterns, and beyond twelve you are usually wasting time once the pattern is clear.

When should I not use pre-build interview validation?▼

Do not use it when a spec is already in engineering review, for ongoing discovery work, or for post-launch validation. Those cases call for spec review, broader discovery, or surveys and analytics instead.

What are JTBD Forces of Progress in user research?▼

Forces of Progress are four pressures determining adoption: push from the current situation, pull toward a new solution, anxiety about switching, and habit keeping users in place. A feature succeeds only when push plus pull clearly exceeds anxiety plus habit across interviewees.

Why do user interviews falsely confirm feature ideas?▼

Interviews falsely confirm ideas when the interviewer pitches the feature, asks hypothetical questions, or has no written failure threshold. Users politely agree, compliments get mistaken for demand, and the team ships a feature nobody switches to.