Working with AI
Learn how to use MCP servers and skills to give AI the right context, connect external tools, and build faster, more reliable workflows in your apps.
Intent UI Skills
Intent UI Skills give AI the context it needs to understand your code, pick the right components, and use the correct APIs, so you can build faster and with more confidence.
npx skills add intentui/skills
How to use it
Using it is simple. In Claude Code, run /intentui. In Codex, run $intentui, then describe what you want to build. You can see a preview here.
MCP
MCP is like a universal plug for AI. It lets an AI connect to tools and data sources (files, apps, APIs) through one standard interface, so it can use them when you ask. Learn more in the MCP documentation.
Do you need MCP
Yes, if you work with AI. MCP is highly recommended because it helps your editor or assistant install components faster and use the correct APIs. If you do not use AI in your workflow, you can ignore it.
shadcn
All registries in Design Intent UI can be installed through the shadcn cli, which is why you can use the shadcn MCP server here. Before you begin, complete the steps in the shadcn CLI section, including setting up your token and configuring your components.json file.
To add components directly from your code editor using the MCP server, choose one of the supported clients below.
npx shadcn@latest mcp init --client codex
Add the following to ~/.codex/config.toml:
React Aria
Most components and design blocks in Design Intent UI are built on top of React Aria. To help AI use the correct React Aria APIs and patterns, the React Aria MCP server is strongly recommended.
codex mcp add react-aria -- npx @react-aria/mcp@latest
Add the following to ~/.codex/config.toml: