refero-design

Ground UI design decisions in real design research using Refero MCP tools.

2|Updated Feb 4, 2026
One-click install
npx skills add https://github.com/bbssppllvv/refero_skill --skill refero-design-bbssppllvv
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: refero-design
Source: https://github.com/bbssppllvv/refero_skill/tree/main
Command: npx skills add https://github.com/bbssppllvv/refero_skill --skill refero-design-bbssppllvv

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Ground decisions in real design research rather than generic AI output, reducing guesswork and increasing design confidence.

Core Features & Use Cases

  • Research-first workflow guiding discovery, analysis, and design using Refero MCP and tools (search_screens, search_flows, get_screen, get_flow, get_design_guidance).
  • Pattern extraction from real products to inform typography, color, layout, and interaction choices.
  • Anti AI-slop focus to ensure intentional design, credible craft, and unique brand voice.
  • Deliverables include research summaries, steal lists, and production-ready design specifications.

Quick Start

Install the skill and run the example workflow to collect references using search_screens and get_screen.

Frequently Asked Questions about refero-design

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

FAQPage Schema
How do I ground UI design decisions in real design research instead of generic AI output?

Grounding UI design decisions in real design research involves searching and analyzing references from existing products using MCP tools. This approach extracts actual patterns for typography, color, and layout, producing a research summary and production-ready design specification to reduce guesswork.

What's the best way to research design patterns for new screens and flows?

Researching design patterns for new screens and flows uses MCP tools like search_screens and search_flows to discover real product examples. You then analyze these references with get_screen and get_flow to extract intentional design choices, generating a steal list for your interface.

Do I need the Refero MCP tools to use this research-first design workflow?

Yes, the research-first design workflow requires using Refero MCP tools such as search_screens, search_flows, get_screen, get_flow, and get_design_guidance. These tools enable the discovery, analysis, and pattern extraction needed to produce credible craft and avoid generic AI outputs.

Can I create production-ready design specifications for new interfaces using this approach?

Yes, you can create production-ready design specifications by applying a research-first workflow that discovers and analyzes real interface references. The process yields a research summary, extracted patterns, and a final specification grounded in actual design research rather than AI guesswork.

How does design research with MCP tools compare to generating UI from generic AI?

Design research with MCP tools pulls patterns from real products to inform typography, color, and interaction choices, ensuring intentional design and unique brand voice. Generic AI output lacks this research grounding, often producing guesswork and recognizable AI-slop.

When should I not use research-first design for my UI tasks?

You should not use research-first design when you need rapid, low-fidelity prototyping without referencing existing products, or if your interface requires highly novel interactions that cannot be informed by analyzing current screen and flow patterns from other apps.