Token导航 LogoToken导航TokenDH.com
前端设计external-servicegithub未标认证来源可访问许可证需确认审计提醒

figma-mcpFigma MCP 浏览器

Agent Skill

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

总安装

1,523

周安装

61

GitHub Stars

22

下载量

493
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:figma-mcp(Figma MCP 浏览器)
来源仓库:https://github.com/tdimino/claude-code-minoan
仓库路径:skills/figma-mcp
安装命令:
npx skills add https://github.com/tdimino/claude-code-minoan --skill figma-mcp
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/tdimino/claude-code-minoan --skill figma-mcp

简介

用于优化界面设计、视觉规范、排版、配色、布局和交互体验。

  • 可整理页面结构、生成 UI 方案或检查视觉一致性,改进组件层级。
  • 需结合品牌、设计系统和用户任务,避免堆砌装饰元素;改动后应通过截图或浏览器预览检查表现。
  • 安装命令:npx skills add https://github.com/tdimino/claude-code-minoan --skill figma-mcp
  • 适用于 Codex、Claude、Cursor、Gemini CLI,需授权访问 Figma 文件。

SKILL.md

Figma MCP

Overview

This skill enables accurate design-to-code conversion by leveraging Figma's MCP (Model Context Protocol) server to access structured design data directly from Figma files. Unlike screenshot-based approaches, the Figma MCP provides semantic information about every design element including exact spacing, colors, typography, component hierarchy, and design system tokens, resulting in significantly more accurate code generation.

When to Use This Skill

Trigger this skill when users:

  • Provide Figma URLs (file links or frame selection links)
  • Request converting/implementing Figma designs into code
  • Ask to "build this from Figma" or "implement this design"
  • Need to extract design tokens, variables, or design system values
  • Want to update existing code to match a Figma design
  • Mention maintaining design-to-code consistency

Workflow

Step 1: Receive Figma Design Reference

When a user provides a Figma URL or requests design implementation:

  1. Identify the Figma link type:

- File URL: https://www.figma.com/file/[FILE_KEY]/... - Frame selection: https://www.figma.com/file/[FILE_KEY]/...?node-id=[NODE_ID] - Design URL: https://www.figma.com/design/[FILE_KEY]/...

  1. Confirm access:

- Verify the Figma MCP server is configured and accessible - If not configured, reference references/setup-guide.md for setup instructions

Step 2: Fetch Design Data Using MCP Tools

Use the available Figma MCP tools to retrieve structured design data:

Key MCP Tools Available:

  • get_figma_data - Fetch complete design data for a file or specific frame
  • get_figma_variables - Extract design tokens (colors, spacing, typography)
  • get_figma_components - Retrieve component definitions and instances
  • get_figma_styles - Access text, color, and effect styles

Best Practice: Start with get_figma_data for the specific frame or node the user referenced. This provides the complete structure, layout data, and styling information.

Example workflow:

User: "Implement this Figma design: [Figma URL with node-id]"

1. Extract file_key and node_id from URL
2. Call get_figma_data with file_key and node_id parameters
3. Receive structured JSON with:
   - Element hierarchy (frames, groups, components)
   - Layout properties (position, size, constraints, auto-layout)
   - Styling (fills, strokes, effects, typography)
   - Component instances and variants
   - Design tokens and variables

Step 3: Analyze Design Structure

Before generating code, analyze the fetched design data:

  1. Identify component hierarchy:

- What are the main containers/sections? - Which elements are reusable components? - How does the layout flow (flex, grid, absolute)?

  1. Extract design system values:

- Colors (fills, strokes, backgrounds) - Typography (font families, sizes, weights, line heights) - Spacing (padding, margins, gaps) - Borders, shadows, effects

  1. Determine layout approach:

- Auto-layout frames → Flexbox or CSS Grid - Absolute positioning → CSS absolute/relative positioning - Constraints → Responsive behavior

Step 4: Generate Code

Generate production-ready code based on the design data:

Code generation principles:

  1. Use semantic HTML:

- Frames → <div>, <section>, <header>, etc. - Text → <h1>-<h6>, <p>, <span> - Buttons → <button> - Images → <img>

  1. Preserve design system values:

- Extract color variables from Figma variables - Create CSS custom properties for spacing, colors, typography - Use exact values from design (don't approximate)

  1. Match layout behavior:

- Auto-layout with padding/gap → Flexbox with gap/padding - Auto-layout direction → flex-direction - Constraints → Responsive CSS (flex-grow, min/max-width)

  1. Component-based architecture:

- Figma components → React/Vue components or CSS classes - Component instances → Component usage with props - Variants → Props or conditional classes

  1. Responsive considerations:

- Use relative units where appropriate (rem, em, %) - Consider breakpoints for desktop/mobile variants - Preserve constraint behavior from Figma

Example output structure:

// For React
import React from 'react';
import './styles.css';

export const ComponentName = () => {
  return (
    <div className="container">
      {/* Preserved hierarchy and styling */}
    </div>
  );
};

Step 5: Validate and Refine

After generating initial code:

  1. Verify accuracy:

- Do spacing values match the design exactly? - Are colors accurate (check hex values)? - Does the layout behavior match (responsive, wrapping)?

  1. Check for improvements:

- Can any values be extracted to CSS variables? - Are there repeated patterns that should be componentized? - Is the code accessible (semantic HTML, ARIA attributes)?

  1. Offer refinements:

- Suggest design system integration if applicable - Mention any design inconsistencies detected - Propose optimizations (shared styles, component extraction)

Best Practices

Design-to-Code Accuracy

  • Never approximate values: Use exact pixel values, colors, and measurements from Figma data
  • Preserve hierarchy: Maintain the nesting structure from Figma frames/groups
  • Match auto-layout behavior: Figma's auto-layout maps directly to Flexbox properties
  • Use design tokens: Extract and reuse color/spacing variables from Figma Variables

Code Quality

  • Semantic HTML: Use appropriate HTML elements, not just divs
  • Component extraction: Identify reusable components from Figma components
  • CSS organization: Separate layout, typography, and visual styling
  • Accessibility: Add appropriate ARIA labels and semantic structure

Design System Integration

  • Check for variables: Use Figma Variables for design tokens
  • Component consistency: Match Figma component names to code component names
  • Code Connect: If available, reference Code Connect mappings for components
  • Naming conventions: Follow existing project conventions for classes/components

Performance

  • Optimize images: Suggest appropriate image formats and compression
  • CSS efficiency: Avoid redundant styles, use CSS variables
  • Bundle size: Keep generated code lean and modular

For more detailed guidance, reference:

  • references/setup-guide.md - Setting up Figma MCP server
  • references/best-practices.md - Extended best practices and examples

Common Patterns

Pattern 1: Landing Page from Figma

User provides: Figma link to landing page design
Process:
1. Fetch complete page structure with get_figma_data
2. Identify sections (hero, features, CTA, footer)
3. Generate semantic HTML with section elements
4. Extract colors/typography to CSS variables
5. Implement responsive behavior from constraints

Pattern 2: Component Implementation

User provides: Figma link to button component with variants
Process:
1. Fetch component definition with get_figma_components
2. Identify variants (primary, secondary, sizes)
3. Create React component with props for variants
4. Extract shared styles to base component
5. Implement variant-specific styles as props/classes

Pattern 3: Design System Extraction

User requests: Extract design tokens from Figma file
Process:
1. Fetch variables with get_figma_variables
2. Fetch text/color styles with get_figma_styles
3. Generate CSS custom properties or theme object
4. Document token naming and usage
5. Export as stylesheet or config file

Troubleshooting

MCP server not available:

  • Check if Figma MCP server is configured in MCP settings
  • Reference references/setup-guide.md for setup instructions
  • Verify user has appropriate Figma plan (Dev Mode access for official server)

Design data incomplete:

  • Ensure the Figma file is accessible (permissions)
  • Verify the node-id in the URL is correct
  • Try fetching parent frame if child node returns limited data

Code doesn't match design visually:

  • Double-check spacing and sizing values
  • Verify color values are exact hex codes from Figma
  • Check if fonts are loaded correctly
  • Review auto-layout direction and alignment properties

Resources

This skill includes reference documentation to support Figma MCP workflows:

references/setup-guide.md

Complete setup instructions for configuring Figma MCP server (official and community options) in Claude Code and other MCP clients. Reference when users need to set up the integration.

references/best-practices.md

Extended best practices, detailed examples, and advanced patterns for design-to-code conversion including component architecture, responsive design strategies, and design system integration.

Recent Updates

Community server rebranded (Apr 2026): The popular community MCP server by GLips has been rebranded from "figma-developer-mcp" to Framelink MCP for Figma (v0.11.0, 14.5k+ stars). It now supports stateless HTTP transport, VS Code layer generation, and prompt injection defense. The npm package name remains figma-developer-mcp for backward compatibility. See references/setup-guide.md for updated configuration.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude

33.47%
按下载量换算165

Codex

33.26%
按下载量换算164

Cursor

17.28%
按下载量换算85

Gemini CLI

10.17%
按下载量换算50

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills