AETumi MCP exposes 10 read-only tools over the Model Context Protocol (Streamable HTTP, JSON-RPC 2.0). Six are catalog tools for searching the AETumi 3D web library; four are experience-discovery tools that match by design intent — industry, visual style or business goal. Each tool is written so an AI coding agent (Claude Code, Cursor, Codex) can pick the right one from a natural-language request.
claude mcp add --transport http aetumi https://mcp.aetumi.appCatalog tools
search_3d_web_assets
What it does. Searches the AETumi library of premium 3D web assets — Three.js / WebGL hero sections, 3D product viewers, shader backgrounds, scroll animations, React and Next.js components and full website templates.
When to use. The user names what to build or a technology.
Arguments. query, category (template, 3d_scene, section, background, gradient), industry, limit (1-30), cursor.
Example. “Find a Three.js hero for a luxury product website.”
get_3d_web_asset
What it does. Full public detail of one asset by id — category, industry, tags, tech stack, live demo, pricing.
When to use. After a search, to inspect one component or template.
Arguments. id (required).
list_3d_web_categories
What it does. Browses the library by category and industry with counts — heroes, product viewers, shader/WebGL backgrounds, scroll animations, sections, templates across 20+ industries.
When to use. To see what is available before searching.
recommend_3d_web_stack
What it does. Recommends assets plus a production build approach — Three.js / WebGL architecture, framework (React Three Fiber / Next.js) and performance guidance.
When to use. The user describes a build and wants a starting point plus how to build it well.
Arguments. use_case (required), industry, framework.
get_pricing
What it does. AETumi lifetime, buy-once plans and what each includes. When to use. Cost, license or what-you-get questions.
about_aetumi
What it does. Canonical description of AETumi and its links. When to use. “What is AETumi?” / entity questions.
Experience-discovery tools (match by design intent)
These serve designers and product teams as much as developers — they speak the language of industry, style and goal, not Three.js internals.
filter_by_industry
What it does. Finds premium 3D web experiences for an industry — Beauty & Cosmetics, Automotive, Fashion, Real Estate, E-commerce, SaaS, Food & Beverage, Travel, Finance.
Arguments. industry (required), limit.
Example. “Show me 3D web experiences for automotive.”
filter_by_style
What it does. Finds experiences by visual style — dark cinematic, editorial luxury, glass luxury, bright architectural, technical precision.
Example. “A dark cinematic 3D landing page.”
filter_by_goal
What it does. Finds experiences by business goal — product launch, collection launch, premium e-commerce, property presentation, portfolio.
Example. “A premium e-commerce 3D experience.”
find_experiences
What it does. Natural-language search across industry, style and goal at once — the best single entry point when a request mixes all three.
Arguments. query (required), limit.
Example. “dark cinematic automotive launch” · “editorial luxury fashion lookbook”
The intent to tool to result pattern
Every tool answers one thing: intent → tool → result. An AI assistant reads the request, picks the tool whose name and description match the intent, gets a real, ranked result with a live reference, and brings it into the build. You keep the editable source and full control of architecture, performance and accessibility.