Included with aiverse Pro

Your agent can code.Now it can design AI-UX.

aiverse MCP connects Claude, Cursor, and your favorite agents to 31 AI-UX patterns, 30 expert insights, and 314 real-product examples — so what they build starts from what already works.

Install MCPSee plans
The problem

Your agent writes the code. It doesn't know what works.

AI-UX is a new discipline — building trust, streaming output, refinement loops, learning from feedback. Your agent has never studied how the best AI products handle them. So it guesses… and it shows.

aiverse MCP fixes that.

One semantic index. Everything aiverse knows.

Your agent searches it all with plain language — or a screenshot.

31
AI-UX design patterns

Full design guides — when to use them, why they matter, how to build them.

30
insight articles

Deep dives on how real AI products design onboarding, trust, and refinement.

314
real-product videos

Interaction recordings from shipped AI products, tagged by platform and industry.

Ask anything. Get design answers back.

Prompts you can copy straight into Claude Code, Claude Desktop, or Cursor once connected.

Which onboarding patterns work for an AI feature users don't trust yet?
Review this screenshot of my chat UI against aiverse patterns.
Show me real products streaming AI output on mobile.
Plan the UX for an AI summarization feature, phase by phase.
How do the best AI products handle regenerate and edit flows?
Find real examples of multimodal input in design tools.

Five tools. Three guided workflows. One library.

Everything your agent needs to search, read, and reference the aiverse AI-UX knowledge base.

search_aiux_library

Semantic search across every pattern, insight, and video — with plain language or UI screenshots.

list_patterns

Browse all 31 patterns by journey stage: onboarding, input, output, refinement, learning.

get_pattern

Pull the full design guide for any pattern — when to use it, why it matters, design guidance.

get_insight

Read full aiverse insight articles — analyses of how real AI products design their UX.

find_product_examples

Find real-product interaction videos by pattern, platform, or industry.

Guided prompts, ready to run

Type / in Claude Code — or open the + menu in Claude Desktop — and these appear as one-click workflows.

/design_review
Design review

Critique a screen or flow pattern-by-pattern against the aiverse library. Paste a screenshot, get a ranked review.

/plan_ai_feature_ux
Plan an AI feature

A phase-by-phase UX plan for any AI feature — onboarding through learning — with a real product to study per phase.

/find_real_examples
Find real examples

Curated real-product videos of any AI interaction, filtered by platform and industry.

Built for the moments design decisions happen.

Run a design review
  • Paste a screenshot or describe the flow.
  • Your agent checks it against 31 AI-UX patterns.
  • Get a pattern-by-pattern critique, ranked by impact.
Plan an AI feature
  • Describe the feature you're building.
  • Get a phase-by-phase UX plan across the AI journey.
  • Every phase backed by patterns and a real product to study.
Ground it in real products
  • Ask for real examples of any AI interaction.
  • Filter by platform or industry.
  • 314 interaction videos, each linked back to aiverse.design.
"aiverse has significantly eased my process of exploring emerging AI UX patterns, which is why I keep coming back to the platform!"
Niomi Shah
Niomi Shah
Designing GenAI Experiences at Intuit

Set up in under a minute

One sign-in. No API key. Works with Claude, Cursor, and any MCP client.

  1. 1

    Get aiverse Pro

    aiverse MCP is included with every Pro membership — no extra add-on.

    See plans
  2. 2

    Register the aiverse MCP server

    Open a terminal and run this command.

    Terminal
    claude mcp add --transport http aiverse-ai-ux https://api.aiverse.design/mcp
  3. 3

    Authorize

    Start a new Claude Code session and run /mcp. Select aiverse-ai-ux, then Authenticate. A browser window opens — sign in with your aiverse account and approve.

    Claude Code
    /mcp
  4. 4

    You're set

    Start prompting with the aiverse library behind your agent. Try: “Which patterns should my AI onboarding use?”

Questions, answered

Give your agent a design brain

Whatever AI feature you're building, an aiverse pattern has already mapped it. Install the MCP and start from what works.

Install aiverse MCP

Included with Pro — see plans