Why UX designers should care about model context protocol

When the experience goes outside the application

Think of Model Context Protocol as the universal translator for AI systems. If you’ve ever wished your AI assistant could actually access your calendar, update your project management tools, and pull data from your CRM all in one conversation instead of making you jump between seventeen different apps, that’s exactly what MCP enables.

Here’s the thing that got my attention: MCP isn’t just another API. It’s specifically designed for AI agents to communicate with external tools and data sources. Anthropic (the company behind Claude) developed it as an open standard, which means it’s not locked into any single platform. When I say “open standard,” I mean anyone can implement it, which historically has been a good sign for widespread adoption.

What Is MCP, and why Should you care?

What this means for UX designers: We’re moving from designing isolated app experiences to designing intelligent workflows that span multiple systems. Instead of thinking about how users navigate through menus and forms, we’re starting to think about how users express intent and watch AI orchestrate complex tasks behind the scenes.

Designing with MCP Server: Bridging Design Systems and AI for Developer-Friendly Prototypes

The Shift: From apps to workflows

For almost twenty years, we’ve been designing apps. Individual, isolated experiences where users manually switch between tools, lose context, and follow rigid step-by-step processes. We’ve all used hundreds of these, and they all share the same fundamental limitation: they exist in silos.

MCP changes the game by enabling what I’m calling “workflow design.” Instead of designing a project management app, you’re designing how AI can intelligently coordinate between your project management tool, your calendar, your email, your file storage, and whatever else makes sense for the task at hand.

Real example: A user says, “Update the Q4 roadmap based on last week’s customer feedback.” An MCP-enabled system can:

• Pull customer feedback from support systems
• Update project timelines in your project management tool
• Send notifications to stakeholders via Slack
• Generate updated presentations in your preferred format
• Schedule follow-up meetings

All through a single conversation. No app switching, no context loss, no manual copying and pasting between systems.

Why this matters more than you think

I’ve been teaching design for long enough to know that most “revolutionary” technologies aren’t. But MCP feels different for four reasons:

1. It’s About Intelligence, Not Just Interfaces We’re not just making prettier buttons — we’re designing experiences where AI can intelligently orchestrate complex workflows. This requires a fundamentally different approach to UX thinking.

2. Cross-Platform Becomes Actually Seamless When AI agents can connect to any MCP-enabled service, your design decisions impact entire ecosystems, not just individual apps. This is both exciting and terrifying.

3. Real-Time Context Actually Works MCP enables AI to access live data and tools, creating interfaces that are contextually aware and dynamically responsive. No more static mockups that break when real data shows up.

4. The Network Effect Is Real As more tools adopt MCP standards, designers who understand these interaction patterns will have a massive advantage. It’s like understanding responsive design before everyone else caught on.

The Cost of doings things right

Let’s talk about money, because nobody else will. If you’re a freelance designer or working at a small agency, you’re probably wondering if this is worth the time investment. Based on what I’m seeing, the answer is a cautious yes, but with some important caveats.

The upside: Designers who can think in terms of intelligent workflows rather than static interfaces will command higher rates. Companies are starting to realize they need design thinking that goes beyond traditional UX.

The downside: This requires ongoing learning and collaboration with developers in ways that traditional UX work doesn’t. You’ll need to understand technical concepts without becoming a developer yourself.

My recommendation: Start small. Pick one or two tools you use regularly and experiment with MCP-enabled workflows. Don’t try to redesign everything at once.

New design patterns you should know

Working with MCP-enabled experiences has taught me some new patterns that feel genuinely different from traditional UX:

The “Conductor Interface” Think of this as a central hub where users can see and control AI actions across multiple tools. It’s like a dashboard, but one that thinks and acts on behalf of users. The key is providing oversight without micromanagement.

Progressive Disclosure for AI Actions Show high-level results first, then let users dive into specifics from individual tools when needed. Users want to know what happened, but they don’t want to be overwhelmed by every API call.

Intent-Based Navigation Replace traditional menus with goal-oriented prompts and intelligent suggestions. Instead of “Click here to create a project,” it’s “What would you like to accomplish?”

Context-Aware Widgets Design interface elements that adapt based on available data and tools in the user’s environment. This is personalization that actually works because it’s based on real capabilities, not just user preferences.

Getting started (without going crazy)

If you’re ready to start experimenting with MCP-enabled design, here’s my practical advice:

Learn the Fundamentals

• Read the MCP specification at modelcontextprotocol.io (it’s more accessible than you think)
• Understand basic concepts from Anthropic’s documentation
• Follow MCP examples and case studies to see the protocol in action

Start Small and Experiment

• Design workflows that connect just two or three tools initially
• Test user understanding of cross-tool interactions
• Iterate based on how users adapt to agentic experiences

Think Beyond Current Constraints

• Design for a world where any tool can connect to any other tool
• Consider how your product fits into larger AI workflows
• Plan for increasing AI capabilities and autonomy over time

What this means for your career

I’ve been watching the design industry evolve for long enough to recognize when something fundamental is shifting. MCP represents a move toward more connected, intelligent digital experiences, and designers who understand this shift will have significant advantages.

The opportunity: Create more helpful and intelligent experiences, reduce user friction across tool boundaries, enable truly personalized workflows, and build products that become smarter over time.

The challenge: Learn to design for unpredictable AI behavior, balance automation with user control, create trust in complex multi-tool interactions, and maintain usability as systems become increasingly sophisticated.

The reality: Understanding MCP isn’t about becoming a developer — it’s about evolving into a designer who can create experiences that harness the full potential of AI in an interconnected world.

A personal note

I’m not going to pretend this is easy. Learning MCP while maintaining your current workload and staying current with traditional UX best practices is a lot. But I’ve seen what happens when designers ignore fundamental platform shifts, and it’s not pretty. If you like, you can check out a short presentation about this that I built with Claude.

The future belongs to UX designers who can think beyond individual applications and design for intelligent, connected experiences. By starting to learn MCP today, you’re positioning yourself to create tomorrow’s most impactful digital experiences.

And honestly? After spending months frustrated with AI tools that felt like fancy autocomplete, designing experiences that can actually accomplish complex tasks feels like the most exciting work I’ve done in years.

Want to dive deeper into MCP and AI-enabled design? I’ll be writing more about this as I continue experimenting with these tools in my classes. The landscape is changing quickly, but the fundamental principles of good design — understanding user needs, creating clear mental models, and building trust — remain more important than ever.

Additional reading

How Figma is using Model Context Protocol

How MCP is revolutioniaing Agentic AI

Model Context Protocol tutorial

Why UX designers should care about model context protocol was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.

This post first appeared on Read More