Token导航 LogoToken导航

Agent Skill · 前端设计

figma-design-generate(Figma 设计 generate)

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-generate 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,权限为操作浏览器;许可证信息还不明确,第三方扫描有未通过项,按安装量在前端设计类的 10,216 个 Skill 里排第 8,476。

总安装 282周安装 12GitHub Stars 10下载 99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-generate(Figma 设计 generate)
来源仓库:https://github.com/petbrains/mvp-builder
仓库路径:skills/figma-design-generate
安装命令:
npx skills add https://github.com/petbrains/mvp-builder --skill figma-design-generate
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/petbrains/mvp-builder --skill figma-design-generate

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-generate 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-generate

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/petbrains/mvp-builder --skill figma-design-gen…;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

    它需要操作浏览器,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-generate 怎么安装?

在终端执行 npx skills add https://github.com/petbrains/mvp-builder --skill figma-design-generate。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

figma-design-generate 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-generate 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

figma-design-generate 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-generate 用的人多吗?

公开数据是累计安装 282 次,近一周安装 12 次,下载 99 次,GitHub 10 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-generate 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

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-generate

Figma 设计 generate · 本页

28210操作浏览器未明确未通过

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 Generate

Capture UI and send it to Figma as fully editable design layers — real frames, components, auto-layout, and text layers, not flat screenshots.

The tool is remote only, exempt from standard Figma MCP rate limits, and works with Figma Design files.

When to Use

  • User says "push to Figma", "send to Figma", "capture UI", "code to canvas", "sync Figma"
  • After building or implementing a feature — send result to designers for review
  • Syncing a Figma file with the current implementation state
  • Roundtrip workflow: Figma → code → Figma → designer refines → code again

Process

1. Determine the target

Three destination options:

New Figma file — default when no file specified. Created in the user's team or organization drafts. The tool will ask the user which Figma organization or team to use.

Existing Figma Design file — user provides a Figma file URL. Requires edit permissions. If the URL doesn't point to a Figma Design file, a new file is created instead.

Clipboard — user wants to paste layers into a Figma file manually.

2. Call generate_figma_design

Prompt the tool with a clear description of what to capture and where to send it.

Example prompts to the tool:

  • New file: "Start a local server for my app and capture the UI in a new Figma file."
  • Existing file: "Start a local server for my app and capture the UI in <Figma file URL>."
  • Clipboard: "Start a local server for my app and use the Figma MCP server to capture the UI to my clipboard."

If a local server is already running, omit the "Start a local server" part — the tool can infer it, but being specific the first time helps.

The tool handles everything: starts the server if needed, injects the capture script, and opens a browser window. An initial capture happens automatically when the browser window opens.

3. User interacts with the capture toolbar

In the opened browser window, a toolbar appears with two options:

  • Entire screen — captures the full current state of the interface
  • Select element — lets the user pick a specific element to capture

The user controls this toolbar. Wait for them to complete.

4. Finish the capture

  • If capturing to a Figma Design file — user clicks Open file to view the result
  • If capturing to clipboard — user pastes the design layers into any Figma file
  • The agent can also complete the process via a follow-up prompt

Decision Rules

SituationAction
User provides Figma URLUse as existing file target
No target specifiedDefault to new file — tool asks user for org/team
Multiple pages or states neededCall tool again for each — it reuses the same file within the conversation
User says "also capture checkout" after first captureNo need to re-specify the file — tool infers same target
Capturing from a live web app or production sitePrompt the tool to use Playwright to inject the required script
Capturing from local dev serverNo extra steps — tool handles injection directly
User wants to capture responsive variantsResize the viewport in the browser between captures

Tips for Best Results

Be descriptive about what to capture. "Capture the checkout flow" works better than "Capture this" — helps organize frames in Figma.

Capture strategically. Focus on key screens and important states (empty, error, filled) rather than every page. Responsive breakpoints are valuable for designer review.

Sequential captures reuse the target. After the first capture to a file, subsequent prompts like "Also capture the account screen" automatically go to the same file. The tool infers this, or asks for confirmation.

Existing file suggestions. When capturing to an existing file, the tool can suggest recent files the user has created or edited.

Design tokens alignment. If the project uses design tokens aligned with Figma variables, the captured layers will reference correct tokens — makes designer review faster.

平台分布

Codex

34.19%
按下载量换算34

Claude

30.92%
按下载量换算31

Cursor

16.97%
按下载量换算17

Gemini CLI

9.31%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

未通过

Snyk

可疑

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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