ux-lean-ux

Facilitates Lean UX Build-Measure-Learn cycles producing hypothesis backlogs, assumption maps, and experiment designs.

Updated Jul 2, 2026
One-click install
npx skills add https://github.com/geekatron/jerry-claude-plugin --skill ux-lean-ux-geekatron
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ux-lean-ux
Source: https://github.com/geekatron/jerry-claude-plugin/tree/main/skills/ux-lean-ux
Command: npx skills add https://github.com/geekatron/jerry-claude-plugin --skill ux-lean-ux-geekatron

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Design teams often make opinion-based decisions without testing whether changes actually produce desired outcomes. This Skill structures design decisions as testable hypotheses using Jeff Gothelf and Josh Seiden's Lean UX methodology, enabling evidence-based iteration through Build-Measure-Learn cycles for tiny teams of 1-5 people. ## Core Features & Use Cases - Hypothesis-Driven Design: Structures design decisions in the canonical Lean UX format (outcome, users, change, evidence) with ICE scoring (Impact, Confidence, Ease) for prioritization. - Assumption Mapping: Places assumptions into a 4-quadrant risk/knowledge framework (Q1-Q4) with category classification (Value, Usability, Feasibility) and movement tracking across cycles. - MVP Experiment Design: Selects from 7 experiment types (A/B test, fake door, concierge MVP, Wizard of Oz, paper prototype, smoke test, one-question survey) with measurable success criteria. - Validated Learning Log: Documents completed cycles with evidence, VALIDATED/INVALIDATED results, and pivot/persevere/kill decisions. - Use Case: A product team redesigning a checkout flow uses this Skill to generate a hypothesis backlog, map riskiest assumptions, design a fake door test, and document validated learning before handing results to HEART Metrics measurement. ## Quick Start Ask the /user-experience skill to help you test whether a design change will improve conversion, and the ux-orchestrator will route the request to this Lean UX facilitator.

Frequently Asked Questions about ux-lean-ux

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

FAQPage Schema
How do I write a Lean UX hypothesis statement?

Use the canonical Lean UX format: "We believe [outcome] for [users] if [change] because [evidence]." Each hypothesis needs a measurable outcome, a specific user segment, a concrete design change, and evidence-based reasoning, plus a unique ID like HYP-001.

How do I prioritize hypotheses with ICE scoring?

Score each hypothesis on Impact, Confidence, and Ease using 1-10 scales, then compute ICE as the average of the three. Test higher scores first; break ties by preferring hypotheses in higher-risk assumption quadrants (Q1 before Q2, Q4, Q3).

What experiment types does Lean UX support for testing hypotheses?

Seven types are supported: A/B tests, fake door tests, concierge MVPs, Wizard of Oz tests, paper prototypes, smoke tests, and one-question surveys. Selection depends on traffic availability, hypothesis maturity, team resources, and required confidence level.

When should I use Lean UX versus heuristic evaluation?

Use Lean UX for forward-looking hypothesis testing of proposed design changes through experiments. Use heuristic evaluation for backward-looking assessment of existing interfaces against Nielsen's usability heuristics. Lean UX validates future directions; heuristics diagnose current problems.

What are the limitations of AI-facilitated Lean UX?

Single-facilitator AI output lacks cross-functional team perspective diversity and may miss context-specific hypotheses requiring domain expertise. Assumption quadrant placements reflect AI judgment, so high-stakes Q1 assumptions should be reviewed with domain experts before committing engineering resources.