Brand Rules: Keep AI-generated designs on-brand

AI can help your users get from blank canvas to finished design in minutes. But speed isn't worth much if the output doesn't match your users' brand. With Brand Rules, you can create design guardrails that AI must follow when creating designs—and make sure the output it creates is on brand at first pass.

Why brand consistency matters for agentic design

Let's be honest: your users probably don't get excited about AI that creates something fast if they then have to spend a long time fixing the output. At that point, did they actually save any time? Will they even adopt it into their workflow?

Generic output is one of the main sources of real frustration when working with AI. When we surveyed hundreds of email makers about their experience using AI in their workflow, 47% said they don't trust AI output for email marketing, and 27% said that guaranteeing brand-compliant output would make AI more usable for them.

Trust comes from knowing the AI is working within real guardrails, so it hands back on-brand designs your users can use, not something they have to redo. And if your users can trust your tool’s output, they’re more likely to make your agentic workflows a core part of their production process.

Keep AI-generated designs on brand with Brand Rules

With Brand Rules, you can ensure that every design your users create with AI in Beefree matches their design guidelines. A host application can attach a brandRules object to an AI editing session, whether you’re working with Beefree SDK’s MCP Server or our AI Co-Pilot. Brand Rules can give guidance on three key questions:

• What should things look like?
  presets are the core of Brand Rules. Use them to define styles—colors, typography, and more—for buttons, paragraphs, titles, and other content blocks. Set one preset per element, or give the AI guidance on complex design systems (like a primary and secondary button style) with descriptions that tell the AI when to use each one.
  
• What can the AI touch?
  Stop the AI from adding certain block types, editing others, or deleting content it shouldn't remove, like required disclosure content. This is especially useful if your users work in a regulated environment.
  
• How much content fits into a design?
  Limits cap things like the number of rows, columns per row, or total content blocks, keeping designs within the structure your users expect.

Read the developer docs for a full overview → 

Give AI simple design guidelines or codify even the most complex design systems

Let’s look at a simple example of how you can use Brand Rules to give the AI design guidance: 

Say your user's brand guidelines call for all buttons to be bright blue, with slightly rounded corners and white text. You can define your button presets like this:

{
  "brandRules": {
    "presets": {
      "blocks": {
        "button": {
          "brandButton": {
            "description": "The only approved button style.",
            "properties": {
              "backgroundColor": "#1A73E8",
              "color": "#FFFFFF",
              "borderRadius": "4px",
              "fontWeight": "700"
            }
          }
        }
      }
    }
  }
}

Every time the AI adds a button to the design, it will use the brandButton styles you defined. Every button will look exactly the same.

If your users are working with advanced design systems, a single preset per design element might not be enough. For example, while a primary CTA might be blue with white copy, a style guide might also allow for secondary CTAs of a different style, and you'll want the AI to work with those instructions too.

Presets let you codify even the most complex design systems: you can define multiple presets per block type (e.g. primaryCta, secondaryCta), each with a written description that guides exactly how it should be used. The AI reads the descriptions and picks the right styles for the context:

{
  "brandRules": {
    "presets": {
      "blocks": {
        "button": {
          "primaryCta": {
            "description": "Use for the main call-to-action — one per email.",
            "properties": {
              "backgroundColor": "#1A73E8",
              "color": "#FFFFFF",
              "borderRadius": "4px"
            }
          },
          "secondaryCta": {
            "description": "Use for secondary actions like 'Learn more'.",
            "properties": {
              "backgroundColor": "#FFFFFF",
              "color": "#1A73E8",
              "borderTop": "1px solid #1A73E8",
              "borderBottom": "1px solid #1A73E8",
              "borderLeft": "1px solid #1A73E8",
              "borderRight": "1px solid #1A73E8",
              "borderRadius": "4px"
            }
          }
        }
      }
    }
  }
}

Every button the AI adds to a design will still be on-brand, but now it can choose between two presets—your blue primary button and the white secondary one—and it has clear instructions for when to use each button style. 

Using Brand Rules with Beefree SDK’s MCP Server or AI Co-Pilot​

Whether you're using Brand Rules directly with our MCP Server or planning to use it to set guardrails for our ready-made AI Co-Pilot, the underlying schema is exactly the same.

How you pass Brand Rules into the AI workflow differs slightly by product. Read all about it in our docs:

• How to use Brand Rules with the MCP Server: editor-managed session or API-managed session
• How to use Brand Rules with the AI Co-Pilot

Get started with Brand Rules today

Brand Rules are available in Beta starting today. See the docs to learn more about how to use them, and start teaching your AI how to create on-brand designs.

Questions or feedback on Brand Rules? Reach out! We can't wait to hear from you.