Overview
A brand is the heart of UI Rules. It holds a complete design language plus ruleset, and it is the source of truth that your AI coding tools read over MCP. When an AI tool generates UI, it pulls from the brand so the result matches how your product is meant to look, sound, and behave.
What a brand is
A brand bundles everything an AI tool needs to build on-brand UI:
- Your design tokens (color, typography, radius, shadow, spacing, sizing).
- Your component and block vocabulary.
- Your voice and tone for copy.
- Your accessibility and logo-usage rules.
- The named rule groups that explain how to apply all of the above.
One organization can hold multiple brands, so a single workspace can serve, for example, a primary product brand and a separate marketing brand. Each brand is independent and serves its own ruleset.
The brand dashboard
The dashboard is where you shape and review a brand. Its main areas map to the four parts that compose a brand:
Styles
Library
Content
Rules
When you create a brand it starts on the default theme (an oklch neutral palette, a near-black primary, and the Geist and Geist Mono fonts). A dashboard alert prompts you to import your brand styles until you configure them. See Importing styles to replace the defaults with your own.
The default serving unit and source of truth
A brand is served to your AI tools by default and is inherited by every project (connected app) in the organization. The brand is the source of truth: projects serve the brand ruleset untouched, plus their own sparse overrides, and the brand itself is never modified by a project.
A brand's styles, rules, and component vocabulary are exposed to AI tools (Cursor, Claude Code, Codex, v0, Lovable, Replit, Figma Make) over MCP through a connection key. To wire a tool to a brand, see Connect.
Because the brand is inherited everywhere, a change you make to the brand reaches every connected project at once. Use project overrides only when one app needs to differ.
How the four parts compose a brand
Styles, Library, Content, and Rules work together to make one coherent design language:
- Styles define the raw token values an AI tool builds with, and carry General rules groups per category.
- Library defines the component and block vocabulary, the icon set, and the brand assets the tool can use, each with its own rule groups.
- Content defines the voice and tone for any copy the tool writes, from button labels to error messages.
- Rules are the named guidelines, each with text, a Guidance marker (Do, Don't, Caution, or Info), and an Importance (Required, Recommended, or Optional), that tell the tool how to apply everything else.
Together they form the single ruleset that UI Rules serves over MCP.
