Skip to main content
Glama
kenneives

design-token-bridge-mcp

extract_tokens_from_figma_variables

Extract design tokens from Figma variables JSON export to translate design systems across platforms like Material 3, SwiftUI, and CSS.

Instructions

Parse Figma variables export JSON and extract design tokens

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
variablesYesFigma Variables REST API JSON export as a string

Implementation Reference

  • Main handler function that extracts design tokens from Figma Variables JSON. Parses the input, resolves aliases, categorizes variables by type (COLOR, FLOAT), normalizes names, and returns a DesignTokens object with colors, spacing, and radii.
    export function extractTokensFromFigmaVariables(jsonString: string): DesignTokens {
      let data: FigmaVariablesExport;
      try {
        data = JSON.parse(jsonString);
      } catch {
        throw new Error("Invalid JSON: could not parse Figma variables export.");
      }
    
      // Handle both direct format and nested under meta.variables
      const raw = data as any;
      const variables: Record<string, FigmaVariable> =
        data.variables ?? raw.meta?.variables ?? {};
      const collections: Record<string, FigmaVariableCollection> =
        data.variableCollections ?? raw.meta?.variableCollections ?? {};
    
      if (Object.keys(variables).length === 0) {
        throw new Error("No variables found in the Figma export.");
      }
    
      // Find the default mode ID (first collection's default, or first mode)
      let defaultModeId: string | undefined;
      for (const collection of Object.values(collections)) {
        defaultModeId = collection.defaultModeId;
        break;
      }
    
      const tokens: DesignTokens = {};
      const colors: Record<string, ColorToken> = {};
      const spacing: Record<string, number> = {};
      const radii: Record<string, number> = {};
    
      for (const variable of Object.values(variables)) {
        // Resolve the value for the default mode
        const modeId =
          defaultModeId ?? Object.keys(variable.valuesByMode)[0];
        let value = variable.valuesByMode[modeId];
    
        // Resolve aliases
        value = resolveAlias(value, variables, modeId, 10);
    
        // Clean up variable name: "Colors/Primary" → "primary", "Spacing/xs" → "xs"
        const name = normalizeVariableName(variable.name);
    
        switch (variable.resolvedType) {
          case "COLOR": {
            if (isColorValue(value)) {
              const hex = figmaColorToHex(value as { r: number; g: number; b: number; a: number });
              const category = inferColorCategory(variable.name);
              colors[name] = {
                value: hex,
                ...(variable.description ? { description: variable.description } : {}),
                ...(category ? { category } : {}),
              };
            }
            break;
          }
          case "FLOAT": {
            if (typeof value === "number") {
              const lowerName = variable.name.toLowerCase();
              if (lowerName.includes("radius") || lowerName.includes("corner")) {
                radii[name] = value;
              } else {
                // Default FLOAT to spacing
                spacing[name] = value;
              }
            }
            break;
          }
          // STRING and BOOLEAN are not directly mappable to our token schema
        }
      }
    
      if (Object.keys(colors).length > 0) tokens.colors = colors;
      if (Object.keys(spacing).length > 0) tokens.spacing = spacing;
      if (Object.keys(radii).length > 0) tokens.radii = radii;
    
      const hasContent = Object.values(tokens).some((v) => v !== undefined);
      if (!hasContent) {
        throw new Error("Figma variables found but none could be mapped to design tokens.");
      }
    
      return tokens;
    }
  • Output schema defining the universal DesignTokens interface structure with colors, typography, spacing, radii, elevation, and motion properties.
    export interface DesignTokens {
      colors?: Record<string, ColorToken>;
      typography?: Record<string, TypographyToken>;
      spacing?: Record<string, number>; // px values
      radii?: Record<string, number>; // px values
      elevation?: Record<string, ElevationToken>;
      motion?: Record<string, MotionToken>;
    }
  • Input schema defining Figma Variables API response shape including FigmaVariablesExport, FigmaVariable, FigmaVariableCollection, and FigmaValue types for parsing Figma's JSON format.
    interface FigmaVariablesExport {
      variables?: Record<string, FigmaVariable>;
      variableCollections?: Record<string, FigmaVariableCollection>;
    }
    
    interface FigmaVariable {
      id: string;
      name: string;
      resolvedType: "COLOR" | "FLOAT" | "STRING" | "BOOLEAN";
      valuesByMode: Record<string, FigmaValue>;
      description?: string;
    }
    
    interface FigmaVariableCollection {
      id: string;
      name: string;
      modes: { modeId: string; name: string }[];
      defaultModeId: string;
    }
    
    type FigmaValue =
      | { r: number; g: number; b: number; a: number } // COLOR
      | number // FLOAT
      | string // STRING
      | boolean // BOOLEAN
      | { type: "VARIABLE_ALIAS"; id: string }; // alias
  • src/index.ts:75-94 (registration)
    Tool registration with MCP server. Defines the tool name 'extract_tokens_from_figma_variables', description, input schema (variables parameter as string), and async handler that calls extractTokensFromFigmaVariables.
    server.registerTool(
      "extract_tokens_from_figma_variables",
      {
        description:
          "Parse Figma variables export JSON and extract design tokens",
        inputSchema: {
          variables: z
            .string()
            .describe("Figma Variables REST API JSON export as a string"),
        },
      },
      async ({ variables }) => {
        try {
          const tokens = extractTokensFromFigmaVariables(variables);
          return toolResult(JSON.stringify(tokens, null, 2));
        } catch (e) {
          return errorResult(e);
        }
      }
    );
  • Helper functions supporting the main handler: resolveAlias (resolves VARIABLE_ALIAS references), isColorValue (type guard for color values), figmaColorToHex (converts RGB to hex), normalizeVariableName (cleans variable names), and inferColorCategory (categorizes colors by name patterns).
    function resolveAlias(
      value: FigmaValue,
      variables: Record<string, FigmaVariable>,
      modeId: string,
      maxDepth: number
    ): FigmaValue {
      if (maxDepth <= 0) return value;
      if (
        typeof value === "object" &&
        value !== null &&
        "type" in value &&
        value.type === "VARIABLE_ALIAS"
      ) {
        const target = variables[value.id];
        if (target) {
          const targetValue =
            target.valuesByMode[modeId] ?? Object.values(target.valuesByMode)[0];
          return resolveAlias(targetValue, variables, modeId, maxDepth - 1);
        }
      }
      return value;
    }
    
    function isColorValue(value: unknown): boolean {
      return (
        typeof value === "object" &&
        value !== null &&
        "r" in value &&
        "g" in value &&
        "b" in value
      );
    }
    
    function figmaColorToHex(color: { r: number; g: number; b: number; a?: number }): string {
      // Figma colors are 0-1 float range
      const r = Math.round(color.r * 255)
        .toString(16)
        .padStart(2, "0");
      const g = Math.round(color.g * 255)
        .toString(16)
        .padStart(2, "0");
      const b = Math.round(color.b * 255)
        .toString(16)
        .padStart(2, "0");
      return `#${r}${g}${b}`.toUpperCase();
    }
    
    function normalizeVariableName(name: string): string {
      return name
        .split("/")
        .pop()! // Take last segment: "Colors/Primary" → "Primary"
        .replace(/\s+/g, "-") // Spaces to hyphens
        .replace(/([a-z])([A-Z])/g, "$1-$2") // camelCase to kebab
        .toLowerCase();
    }
    
    function inferColorCategory(
      name: string
    ): ColorToken["category"] | undefined {
      const lower = name.toLowerCase();
      if (lower.includes("primary")) return "primary";
      if (lower.includes("secondary")) return "secondary";
      if (lower.includes("tertiary")) return "tertiary";
      if (lower.includes("neutral") || lower.includes("gray") || lower.includes("grey"))
        return "neutral";
      if (lower.includes("error") || lower.includes("danger") || lower.includes("destructive"))
        return "error";
      if (lower.includes("surface")) return "surface";
      if (lower.includes("background") || lower.includes("bg")) return "background";
      return undefined;
    }
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. It states the tool parses and extracts tokens but doesn't describe what 'extract' entails (e.g., format of output, whether it transforms data, error handling, or performance considerations). For a tool with no annotation coverage, this leaves significant gaps in understanding its behavior beyond the basic purpose.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that directly states the tool's purpose without unnecessary words. It is front-loaded with the core action ('parse' and 'extract'), making it easy to scan and understand quickly. Every part of the sentence contributes essential information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's moderate complexity (parsing JSON to extract tokens), no annotations, and no output schema, the description is minimally adequate but incomplete. It covers the basic purpose and input but lacks details on output format, error cases, or how it differs from siblings. For a tool with no structured behavioral data, it should provide more context to be fully helpful.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, clearly documenting the 'variables' parameter as 'Figma Variables REST API JSON export as a string'. The description adds no additional semantic context beyond this (e.g., example JSON structure, validation rules, or common pitfalls). With high schema coverage, the baseline score of 3 is appropriate as the schema does the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the specific action ('parse' and 'extract'), the resource ('Figma variables export JSON'), and the output ('design tokens'). It distinguishes itself from siblings like 'extract_tokens_from_css' or 'extract_tokens_from_json' by specifying the Figma source format, making the purpose unambiguous and well-differentiated.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives. It doesn't mention prerequisites (e.g., needing Figma variables data), exclusions (e.g., not for CSS files), or comparisons to siblings like 'extract_tokens_from_json' (which might handle generic JSON). Without such context, the agent must infer usage from the tool name alone.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/kenneives/design-token-bridge-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server