SkillFoxx

Gravity AI UI MCP

An interface generator that turns a model's description into a validated React component built on the Gravity UI library

  • 1 ★ on the repository
  • Last commit
  • Apache-2.0
MCP server

Extended access

Why this level

  • The remote MCP server forwards the user request to OpenAI
  • The model output is rendered only through Gravity UI components, no raw HTML
All reasons and checks

astandrik/gravity-ai-uiAuthor:astandrik

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add mcp/gravity-ai-ui

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The remote MCP server forwards the user request to OpenAI
  • The model output is rendered only through Gravity UI components, no raw HTML

This is third-party code. Review the repository files before installing.

What it does

The app takes a user request, asks an OpenAI model to describe an interface in a strict structured A2UI format, and renders the result only with trusted Gravity UI components: cards, forms, tables, metrics, alerts and navigation. The model never emits raw HTML or React directly, only a data description that the app turns into components itself. There is a chat with history, an inspector with preview, React code, A2UI and payload tabs, and copyable React code for the resulting interface. Liked results can be saved and reused as examples for later generations.

Who it is for. For developers who want a working Gravity UI interface prototype quickly from a text description.

Good fit when

  • You need a quick screen prototype in Gravity UI components from a text prompt
  • You need ready React code to refine by hand afterwards
  • You need an MCP tool to search and generate saved interface examples

Not a fit when

  • You need a production-ready interface with no manual work
  • You have no OpenAI key: generation only works through it

Example request

Generate a user profile card with an avatar, activity metrics and an edit button using Gravity UI components

MCP

Transport
http
Authentication
not required

Environment variables

OPENAI_API_KEY required, secret
OpenAI key for interface generation
YDB_CONNECTION_STRING secret
YDB connection string for saving liked interfaces

Limitations

Interface generation runs through the OpenAI API and needs its key; there is no Russian LLM in the project. YDB is only for saving liked results, and basic generation works fine without it. This is not an MCP server for cloud management: the remote MCP only searches, fetches, generates and refines interfaces, not Yandex Cloud resource operations.

How to disable. Remove the gravity-ai-ui MCP connector from your client settings, or simply stop using the web app.

FAQ

What provider powers generation?

The OpenAI Responses API; your own key is passed via OPENAI_API_KEY when running locally.

Can I get ready React code?

Yes, the React tab in the inspector and the get_interface MCP tool return copyable code for the resulting component.

Discussion

No questions or reviews yet.

MCP servers

dsh-generative-ui turns a DeepSeek Harness reply into a live React interface that renders while the model is still typing it.

PluginBroad access206

A DeepSeek Harness plugin where the agent builds a task-specific React app and reads the user's saved choices in the next turn.

PluginExtended access127
Official

MCP server of a React component library for agent interfaces: the agent finds a component and adds it to your project

MCP serverExtended access750

A skill that bans the stock AI interface with floating cards, gradients and glass and sets plain sizes, radii and shadows like Linear

SkillMinimal access2.7K

A Codex skill designs web interfaces in one of five industrial styles at four levels of depth

SkillExtended access295

Finds AI-look tells in generated frontend, from purple gradients to cream and terracotta, and rewrites the interface

SkillExtended access161

README in short

The README describes a Next.js 16 app on React 19 and Gravity UI that generates interfaces via the OpenAI Responses API, with optional storage of liked results in YDB. It lists agent-facing resources: llms.txt, OpenAPI, an MCP server card and comparison pages against other interface generators. It separately covers OpenAI and YDB environment variables and development commands.

README badge

Report a problem