For AI Assistants (Robots)

Atlas ships with built-in resources so AI coding assistants can understand your components, props, and usage patterns out of the box.

LLM Endpoints

The documentation site exposes machine-readable endpoints that you can feed directly to any LLM or AI coding assistant.

MCP Server

Atlas publishes an MCP (Model Context Protocol) server as part of the @pros/atlas package. This gives your AI coding assistant direct access to component documentation, props, and usage examples without leaving your editor.

The MCP server provides three tools:

• list-all-documentation — lists all available components and their metadata
• get-documentation — retrieves detailed documentation for a specific component
• get-documentation-for-story — retrieves documentation for a specific Storybook story

Configuration

Add the MCP server to your editor or AI tool configuration. There are two options depending on whether you want to use the version installed in your project or always fetch the latest.

Project-pinned version (recommended)

Uses the exact version of @pros/atlas installed in your project, so docs always match your components:

{
  "servers": {
    "atlas": {
      "command": "pnpm",
      "args": ["exec", "@pros/atlas"]
    }
  }
}

{
  "mcpServers": {
    "atlas": {
      "command": "pnpm",
      "args": ["exec", "@pros/atlas"]
    }
  }
}

{
  "mcpServers": {
    "atlas": {
      "command": "pnpm",
      "args": ["exec", "@pros/atlas"]
    }
  }
}

Skills

Coming soon: Pre-built "skills" for popular AI coding assistants like GitHub Copilot, Amazon CodeWhisperer, and Anthropic's Claude Code that you can enable with a single click to give them native understanding of Atlas components and documentation. Stay tuned!