ux-design-pipeline

Orchestrates wireframing, multi-critic UX review, component implementation, and browser testing for UI screens.

Updated Jan 29, 2026
One-click install
npx skills add https://github.com/sparkst/sparkry-claude-skills --skill ux-design-pipeline-sparkst
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ux-design-pipeline
Source: https://github.com/sparkst/sparkry-claude-skills/tree/main/plugins/qdesign/skills/qdesign
Command: npx skills add https://github.com/sparkst/sparkry-claude-skills --skill ux-design-pipeline-sparkst

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Building UI screens often skips usability validation: mockups go straight to code, accessibility issues surface late, and browser flows break silently. This Skill enforces a structured design-to-code pipeline where every screen is wireframed, critiqued by three independent reviewers, implemented, and verified in a real browser before shipping. ## Core Features & Use Cases - Wireframe and Critique Loop: Generates mockups with Stitch, then runs three critics (UX Knowledge, UI Expert, UI-UX Pro) that produce severity-rated findings from P0 blockers to P3 polish items. - Production Component Implementation: Converts approved mockups into working components via Magic MCP (21st.dev) and wires them into application code. - Automated Test-and-Fix Loop: Runs Playwright browser tests plus critic re-reviews on the coded version, fixing P0/P1 findings recursively for up to 4 iterations. - Use Case: Ask for an activity dashboard for project managers aged 35-55 on laptops; the pipeline produces mockups, pauses for your approval at each gate, implements the components, and verifies every user flow in a browser. ## Quick Start Run /qdesign with a screen description and audience profile, for example: design an admin dashboard for operations managers who use desktop browsers.

Frequently Asked Questions about ux-design-pipeline

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

FAQPage Schema
How do I design a UI screen with automated UX review?▼

Invoke the pipeline with a screen description and audience profile. It generates Stitch mockups, runs three critic servers for heuristic, audience, and pattern reviews, then implements components with Magic MCP and verifies flows with Playwright.

What MCP servers are required for the UX design pipeline?▼

Six servers are required: stitch for wireframing, ux-knowledge, ui-expert, and ui-ux-pro for critique, magic (21st.dev) for component building, and playwright for browser testing. The preflight phase checks all of them and stops with install commands if any are missing.

How are UX critique findings prioritized and resolved?▼

Findings are classified P0 through P3 by severity. P0, P1, and major P2 items must be fixed; minor P2 items are fixed if trivial; P3 items are skipped. Conflicts resolve toward usability over aesthetics, audience-specific over generic, and accessibility as non-negotiable.

Does the pipeline test the implemented UI in a real browser?▼

Yes. Phase 6 uses Playwright to navigate pages, click interactive elements, fill forms, check console errors, and capture screenshots. Critics re-review the coded version, and a fix loop runs up to 4 iterations before escalating unresolved issues.

When does the design pipeline pause for user approval?▼

It pauses after wireframing (Phase 2), after mockup revision (Phase 4), and after implementation (Phase 5). The test-and-fix loop in Phase 6 runs autonomously without pausing until exit criteria are met or iterations are exhausted.