Eyedropper: Serving Design Systems to AI Agents via MCP

LinkTool

Eyedropper extracts design tokens including colors, typography, and spacing from production code and serves them to AI agents via the Model Context Protocol (MCP). It automatically validates visual code generations against the design system to ensure brand consistency.

Key points

  1. Extracts design tokens such as colors, typography, and spacing from shipped code using a single prompt
  2. Connects to MCP-compliant agents like Cursor and Claude Code via 'npx eyedropper init'
  3. Automated verification catches off-brand styling in generations and suggests correct design tokens
  4. Data export available at any time as DESIGN.md or standard design-tokens JSON
  5. Free tier available for one project and one user; paid tier covers unlimited projects and teams

Why it matters

AI coding assistants frequently create arbitrary styling variations, such as mismatched colors or border radii. Serving design tokens directly via MCP enforces brand rules without requiring developers to repeatedly paste hex codes or styling instructions into prompts.

Analysis

Eyedropper addresses styling drift caused by LLMs by turning design systems into live infrastructure. Rather than relying on static documentation files that easily become outdated, the tool leverages MCP to configure agents like Cursor and Claude Code to verify generations against tokens automatically. The export options for design-tokens JSON and Markdown ensure teams retain ownership of their design assets.

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Similar in content, found by the AI search.

  • Video

    Video:Build Great Products

    Eyedropper: Design Systems and Automated UI Verification for AI Agents

    Chris from Build Great Products introduces Eyedropper, a cloud tool designed to extract design systems from codebases, URLs, or image references. Delivered to coding agents via the Model Context Protocol (MCP), it enables an automated verification loop that keeps AI-generated interfaces on-brand.

    AI & LLMs· Demo

  • Link:Railly

    Tinte: Compile Design Systems into Agent Plugins

    Tinte is a CLI tool that compiles design systems into machine-readable plugins for coding agents such as Claude Code, Cursor, and Codex. Based on a single configuration file, it generates token APIs, LLM composition rules, and matching themes for editors and terminals.

    AI & LLMs· Tool

  • Link:Benji Taylor, Dennis Jin, Alex Vanderzon

    Agentation: Turning Visual UI Feedback into Structured Context for AI Agents

    Agentation converts in-browser UI annotations into structured context designed for AI coding agents such as Claude Code and Cursor. The tool extracts CSS selectors, source file paths, component hierarchies, and computed styles via clipboard or Model Context Protocol (MCP).

    AI & LLMs· Tool

Let's talk about your project

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.