Token导航 LogoToken导航

Agent Skill · 前端设计

figma-design-tool-workflows(Figma 设计 tool workflows)

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。

figma-design-tool-workflows 是前端设计类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为调用外部服务;许可证信息还不明确,第三方扫描有提醒项,按安装量在前端设计类的 10,216 个 Skill 里排第 9,526。

总安装 212周安装 9GitHub Stars 3下载 74核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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-design-tool-workflows(Figma 设计 tool workflows)
来源仓库:https://github.com/phazurlabs/ux-ui-mastery
仓库路径:skills/figma-design-tool-workflows
安装命令:
npx skills add https://github.com/phazurlabs/ux-ui-mastery --skill figma-design-tool-workflows
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/phazurlabs/ux-ui-mastery --skill figma-design-tool-workflows

这个 Skill 是做什么的

前端设计类 Skill 关注页面结构、视觉规范和组件写法,输出通常是可运行的界面代码。

用于辅助界面设计、视觉规范、排版、配色和布局优化。适合整理页面结构、生成 UI 方案或检查视觉一致性。使用时需结合现有品牌、设计系统和用户任务。涉及真实页面改动时应通过截图或浏览器预览检查表现。figma-design-tool-workflows 属于前端设计类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/figma-design-tool-workflows

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/phazurlabs/ux-ui-mastery --skill figma-design-…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要调用外部服务,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

figma-design-tool-workflows 怎么安装?

在终端执行 npx skills add https://github.com/phazurlabs/ux-ui-mastery --skill figma-design-tool-workf…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

figma-design-tool-workflows 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

figma-design-tool-workflows 需要什么权限?

权限级别是调用外部服务。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。

figma-design-tool-workflows 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

figma-design-tool-workflows 用的人多吗?

公开数据是累计安装 212 次,近一周安装 9 次,下载 74 次,GitHub 3 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

figma-design-tool-workflows 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理前端设计类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

前端设计类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
frontend-design

前端设计 · Codex、Claude

8,823,360126,072只读MIT通过
vercel-composition-patterns

Vercel 组合模式 · Codex、Claude

3,858,81625,893需要联网MIT通过
vercel-react-native-skills

Vercel React native skills 前端 · Codex、Claude

2,468,66425,878只读MIT通过
adapt

adapt 搜索 · Codex、Claude

2,027,04023,125操作浏览器未明确通过
bolder

bolder 分析 · Codex、Claude

2,004,28823,130只读未明确通过
figma-design-tool-workflows

Figma 设计 tool workflows · 本页

2123调用外部服务未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Figma Design Tool Workflows — From Canvas to Production Code

Figma as the Design System of Record

Figma is not a drawing tool. It is the single source of truth for design intent, component architecture, token definitions, interaction specifications, and handoff documentation. When used properly, a Figma file becomes a living specification that machines can read, humans can interpret, and CI/CD pipelines can validate against.

The shift in thinking is fundamental: every layer name, every component property, every variable definition in Figma is data. That data flows downstream into code generation, token pipelines, accessibility audits, and visual regression testing. A sloppy Figma file produces sloppy code. A well-architected Figma file produces components that are correct on the first pass.

This skill covers the full Figma workflow from component architecture through AI-assisted code generation and automated validation.

Component Architecture in Figma

Variants and Properties

Figma components support four property types that map directly to code component APIs:

  • Boolean properties control visibility of child layers. A button's showIcon boolean maps to a React showIcon?: boolean prop. The icon layer is present or absent based on this toggle, which means the generated code conditionally renders the icon.
  • Text properties expose editable text content. A button's label text property maps to a children: string or label: string prop. The designer controls the default, the developer overrides at the call site.
  • Instance swap properties define composition slots. A button's icon instance swap maps to an icon?: ReactNode prop. The slot accepts any icon component from the icon library, and the generated code renders the passed component.
  • Preferred values constrain instance swap choices to a curated set. Rather than showing every component in the file, preferred values limit the swap picker to relevant options (icon library components only, for example).

Variant Architecture

Use variants when a component has visually distinct states that change multiple properties simultaneously. A Button component might have variants across two axes: variant (primary, secondary, ghost, danger) and size (small, medium, large). Each combination is a distinct visual design within the same component set.

Use component properties (booleans, text, instance swaps) when a single property toggles independently. Whether a button shows an icon is independent of its variant and size. Making this a boolean property rather than doubling the variant count keeps the component manageable.

Use separate components when two patterns share no structural DNA. A Button and a Card are separate components, not variants of each other.

Nested Components and Slot Patterns

Production-grade Figma components use nested sub-components to create composition points. A Card component contains a CardHeader sub-component, a CardBody sub-component, and a CardFooter sub-component. Each sub-component is a separate Figma component that the Card references via instance swap slots. This mirrors the compound component pattern in React (<Card><Card.Header /><Card.Body /><Card.Footer /></Card>).

Naming Conventions

Follow a Category/Type/State naming hierarchy:

  • Button/Primary/Default
  • Button/Primary/Hover
  • Button/Secondary/Disabled
  • Input/Text/Default
  • Input/Text/Error
  • Card/Elevated/Default

This naming convention creates automatic grouping in the Figma assets panel and maps cleanly to code organization: the Button directory contains Primary and Secondary variant files, each handling their own states.

Variable Modes

Figma variables replace static styles with dynamic, mode-switchable tokens. Variables organize into collections with multiple modes.

Token Layer Architecture

Structure variable collections in three tiers:

  1. Primitive collection — raw values: blue-500: #3b82f6, spacing-4: 16, radius-md: 8. A single mode; these values never change.
  2. Semantic collection — contextual meaning with modes for light/dark/high-contrast: color-bg-primary resolves to white in light mode and neutral-900 in dark mode. color-text-primary resolves to neutral-900 in light mode and neutral-50 in dark mode.
  3. Component collection — component-specific bindings: button-primary-bg references color-action-primary, which references blue-500 in the primitive layer. Modes here handle density (compact/default/comfortable) or brand switching (brand-a/brand-b).

Mode Switching

Variable modes enable designers to preview every combination without duplicating frames. A single button component, bound to semantic variables, previews as light primary, dark primary, high-contrast primary, light compact, dark comfortable, and every other permutation — all by switching the mode on the enclosing frame.

This is the design equivalent of CSS custom property theming. The structure must match: Figma variable names should correspond directly to CSS custom property names or token identifiers in the codebase.

Auto Layout

Auto Layout is the mechanism that makes Figma components developer-friendly. A component without Auto Layout is a static drawing. A component with proper Auto Layout is a responsive specification that communicates padding, spacing, alignment, and resize behavior to developers.

Key principles:

  • Every component uses Auto Layout. No exceptions. If a component cannot be built with Auto Layout, the design needs to be reconsidered.
  • Padding and spacing use variables. Bind Auto Layout padding and item spacing to spacing variables so that density modes work automatically.
  • Fill container for responsive elements. Text blocks and content areas use "Fill container" width so they stretch to available space, matching CSS flex: 1 behavior.
  • Hug contents for intrinsic sizing. Icons, badges, and labels use "Hug contents" so they size to their content, matching CSS width: auto behavior.
  • Absolute positioning sparingly. Badges positioned on avatar corners and floating action buttons use absolute positioning within Auto Layout. This maps to CSS position: absolute within a position: relative container.

Dev Mode

Figma Dev Mode is the bridge between design and development. When a developer inspects a component in Dev Mode, they see:

  • Component properties mapped to code props
  • CSS values for every selected layer (spacing, colors, typography)
  • Variable references showing the token name rather than the raw value
  • Annotations added by designers explaining behavior, interaction rules, and edge cases
  • Code snippets generated from component configuration
  • Measurements between elements (spacing, alignment guides)

Dev Mode effectiveness depends entirely on how well the designer structured the component. Proper naming, variable binding, and Auto Layout produce clean Dev Mode output. Poor structure produces noise.

Figma MCP — The AI-Assisted Game Changer

The Figma Model Context Protocol (MCP) server allows AI assistants like Claude to read Figma design files programmatically. This is the single most impactful development in design-to-code workflow since component libraries.

With Figma MCP, Claude can:

  • Read the complete structure of a design file (pages, frames, components)
  • Extract component properties, variants, and their values
  • Read variable definitions and mode configurations
  • Understand Auto Layout specifications (padding, spacing, alignment, sizing)
  • Extract color, typography, and effect styles
  • Read prototype flows and interaction definitions

This means a designer can build a component in Figma, and Claude can read that component's specification, generate production-quality code (React, SwiftUI, CSS), and produce accompanying tests — all without a human translating the design spec into a ticket.

The Design-to-Code Flywheel

The flywheel operates in six steps:

  1. Design — Designer builds a component in Figma with proper architecture: Auto Layout, variables, component properties, correct naming.
  2. Read — Claude reads the component specification via Figma MCP: layout structure, token bindings, property API, responsive behavior.
  3. Generate — Claude produces code: React component with TypeScript types, CSS Module with token references, Storybook story with all variants, unit test covering props and accessibility.
  4. Validate — Visual regression testing (Chromatic, Percy) compares the rendered component against the Figma source. Accessibility testing (axe-core) validates WCAG compliance.
  5. Iterate — Designer updates the component in Figma. The MCP read detects changes. Claude generates an incremental code update. Visual regression catches any drift.
  6. Ship — CI/CD pipeline deploys validated components. Token pipeline syncs updated variables. Consuming applications pull the latest version.

Each cycle gets faster. The first pass through the flywheel for a new component might save 50% of development time compared to manual implementation. Subsequent iterations on existing components save 70-80% because the structure, tests, and stories already exist and only need updating.

Cross-References

  • Component code output and patterns — reference component-patterns-code for React, SwiftUI, and CSS component architecture, compound components, render props, and polymorphic patterns that Figma MCP-generated code should follow.
  • Design system governance — reference design-systems-architecture for token hierarchy (primitive, semantic, component), versioning strategy, contribution models, and change management processes that govern the Figma library.
  • Visual design tokens — reference ui-visual-design-system for color theory, typography scale, spacing system, and elevation definitions that underpin Figma variable collections.
  • Accessibility in Figma — reference accessibility-inclusive-design for Figma accessibility plugins (Stark, A11y Annotation Kit), color contrast checking, focus order annotation, and screen reader text specification within design files.
  • Interaction and motion — reference interaction-motion-design for Figma prototyping best practices, smart animate specifications, and motion token definitions that map to CSS transitions and animations.

Key Sources

  • Figma official documentation: Components, Variables, Auto Layout, Dev Mode, Code Connect
  • Figma MCP server documentation and API reference
  • Style Dictionary documentation for multi-platform token transformation
  • Tokens Studio (Figma plugin) documentation for token management
  • Chromatic documentation for visual regression testing with Storybook
  • W3C Design Tokens Community Group specification for token format

平台分布

Codex

34.25%
按下载量换算25

Claude

31.91%
按下载量换算24

Cursor

16.94%
按下载量换算13

Gemini CLI

9.99%
按下载量换算7

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

调用外部服务

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills