lockstep

Design Tokens

Design tokens are named values for color, spacing, and type, shared by design tools and code. The DTCG format reached its first stable version in 2025.

Design tokens are named values for the visual decisions in a design system: colors, spacing, type sizes, radii, shadows. Instead of #4F46E5, code uses color.brand.primary. Change the token once, and every place that uses it updates.

The standard format

The Design Tokens Community Group published the first stable version of its format, 2025.10, on 28 October 2025. Tokens are JSON with $value and $type:

{
  "color": {
    "brand": {
      "primary": { "$value": "#4F46E5", "$type": "color" }
    }
  }
}

The group lists support across tools including Figma, Penpot, Sketch, Tokens Studio, and Style Dictionary.

Tokens and AI coding agents

An AI coding agent copies what it sees. Give it a screenshot or a raw design value and it hard-codes the number. Keep the token file in the repository, generate your CSS variables or theme from it, and the agent can use names instead. Figma's MCP server can also return the variables a design uses, so the agent can match a design value to its token. The full setup is in how to give AI coding agents design context.

What are design tokens?
Named values for the visual decisions in a design system, such as color.brand.primary or space.4, stored in a format both design tools and code can read. Code uses the name instead of a raw value like a hex color.
Is there a standard format for design tokens?
Yes. The Design Tokens Community Group published the first stable version of its format, 2025.10, on 28 October 2025. Tokens are JSON with $value and $type fields.
Why do design tokens matter for AI coding agents?
An agent that reads a raw value from a design writes the raw value into code. If tokens are in the repository, the agent can use the token name, so a later brand change updates everywhere at once.

Related