MCPFast / Tools / Refero MCP: Research-first design skill for AI agents
Refero MCP is a design skill for AI agents, leveraging 150K+ real app screens and flows for a research-first approach.
View on GitHub→Refero MCP is a specialized design skill for AI agents, engineered to integrate a research-first methodology into the agent development process. By drawing upon a vast dataset of over 150,000 real-world application screens and user flows, Refero MCP empowers AI agents to generate designs that are grounded in established user experience principles and proven design patterns. This skill is particularly valuable for developers seeking to build AI agents capable of producing functional, intuitive, and user-centric interfaces.
Refero MCP functions as a component within an AI agent's architecture, providing it with the capability to analyze and synthesize information from a comprehensive library of existing application designs. This allows the AI agent to move beyond purely generative design and instead focus on creating solutions informed by extensive user research and established design best practices. The skill enables agents to understand common UI patterns, effective navigation structures, and successful user flow implementations, leading to more robust and user-friendly design outputs.
Refero MCP is intended for AI developers and teams focused on building sophisticated AI agents for design and development tasks. This includes: