原型 · 预览版:收录官方注册中心里的全部服务器,每 30 分钟同步一次。评价和账户功能将在后续阶段推出。
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚

Frame Relay

作者:Joseph-Brendan版本 0.1.0-beta.0i★ –i在你的电脑上运行
注册名称io.github.Joseph-Brendan/frame-relay

安装

概览

作者提供的说明(英文原文)

MCP server for the Frame-Relay design kit: components, tokens and live Figma selections.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

  • 连接方式: 在你的电脑上运行
  • 传输方式: stdio
  • 分类: 设计、图片与音频

安装

选择你使用的 AI 应用,然后复制设置。

安装之前

  • 基础检查:1 项中通过 1 项
  • 在你的电脑上运行,拥有你的用户账户的权限(文件、网络、应用)。
  • 注册中心没有列出需要填写的密钥。

你需要

  • npm已安装 Node.js(用于 npx 命令)

你用的是哪个应用?

在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。

打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json

{
  "mcpServers": {
    "frame-relay": {
      "command": "npx",
      "args": [
        "-y",
        "@frame-relay/cli"
      ]
    }
  }
}

在终端中运行这条命令

claude mcp add frame-relay -- npx -y @frame-relay/cli

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "frame-relay": {
      "command": "npx",
      "args": [
        "-y",
        "@frame-relay/cli"
      ]
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "frame-relay": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@frame-relay/cli"
      ]
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。

你的选择通过 Cookie 保存在这台设备上,无需账户。

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。

它能做什么

它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。

基础检查i

我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日

  • ✓ 通过

    作者与仓库一致

    注册名称已通过 GitHub 账户 Joseph-Brendan 验证,该账户拥有此仓库

  • … 尚未检查

    仍在维护

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    有许可证

    我们还没有从 GitHub 读取这个仓库

  • – 不适用

    端点有响应

    它在你的电脑上运行,没有互联网端点

  • – 不适用

    使用 OAuth 登录

    在你的电脑上运行,改用本地设置

  • … 尚未检查

    工具描述中没有隐藏指令

    它在你的电脑上运行;我们只能读取互联网端点的工具列表

与“设计、图片与音频”中的其他服务器比较i

  • 星标数多于此分类中其他 2782 个服务器的 0%
  • 发布到注册中心的时间比此分类中其他 2782 个服务器的 48% 更新
  • 基础检查结果等于或优于此分类中其他 2782 个服务器的 100%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。

常见问题

Frame Relay 安全吗?

它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。

它需要哪些密钥或代码?

注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端?

它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。

它还在维护吗?

GitHub 仓库最后更新于–,最新版本(0.1.0-beta.0)于29天前进入注册中心。

哪些应用可以使用它?

支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。

有什么可以替代它?

同一分类中有:Figma Context MCP, Claude Real Video, Figma MCP Server

替代选择

同一分类中其他作者的服务器,按星标数排序。

如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

Figma Context MCP

Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.

作者:GLips广泛使用↓ 16.1万/周★ 1.6万5个月前更新

在你的电脑上运行

Claude Real Video

Let any LLM watch a video locally — and search everything it has ever watched.

作者:HUANGCHIHHUNGLeo广泛使用↓ 1106/周★ 22842个月前更新

在你的电脑上运行

Figma MCP Server

The Figma MCP server brings Figma design context directly into your AI workflow.

作者:figma正在增长★ 215213个月前更新

在云端运行✓ mcp.figma.com

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

Frame-Relay

Turn your Figma components into a design kit your AI coding agent follows.

New to this? Follow the step-by-step setup guide for designers.

Status: public beta. Commands below install the beta tag of the CLI. Expect rough edges and tell us what you find with the beta feedback form.

Frame-Relay is for designers who already work with AI coding tools like OpenCode or Antigravity, and for small teams that want their Figma system to stay the source of truth while AI writes the UI. You keep designing in Figma. Your agent gets exact components, tokens and rules instead of guessing from screenshots.

How it works

  1. Design in Figma. Build components the usual way, using the naming contract so exports stay predictable.
  2. Export with the plugin. The Frame-Relay plugin turns your file into a kit: component specs, design tokens, icons and screenshots as plain JSON and image files.
  3. Run sync. One command writes React + Tailwind components, token CSS, rule files for your AI, and MCP client configs into your project.
  4. Build with your AI agent. Your agent reads the kit through the MCP server, then builds screens that match the design. Rules files keep it from inventing raw buttons or hard-coded colors.

Live mode is the optional fifth step: pair Figma with your agent and let it read the layer you have selected. See Live mode. Export mode always works without it.

60-second quickstart

# 1. Create a React + TypeScript app (Vite shown here; any React app works)

npm create vite@latest my-app -- --template react-ts cd my-app

2. Add Tailwind CSS v4

npm install tailwindcss @tailwindcss/vite

Then add the tailwindcss() plugin to vite.config.ts and

@import "tailwindcss"; to src/index.css.

3. In Figma: run the Frame-Relay plugin, open the Export tab, click Export Kit (.zip)

4. Set up Frame-Relay and sync the kit

npx @frame-relay/cli@beta init --yes npx @frame-relay/cli@beta sync --from ~/Downloads/frame-relay-kit-my-file.zip

5. Check the setup

npx @frame-relay/cli@beta doctor

Open the folder in OpenCode (or your AI tool), then ask: "Build a settings page with the kit components." The agent calls the MCP server for the component list, props and tokens, and builds.

Getting the Figma plugin: Frame-Relay is coming to Figma Community. Until it is approved, open Figma desktop, go to Plugins, Development, Import plugin from manifest, and pick packages/plugin/manifest.json from this repo.

Full walkthrough: Getting started. Ideas first: Concepts.

Setup guide for designers

These steps are written for a designer who has never used a terminal. A terminal is a window where you type short commands to your computer. Your AI coding tool has one built in.

Part 1: Get ready (once)

Step 1. Install Node.js. Go to nodejs.org, download the LTS version, and install it like any other app. _Why:_ Frame-Relay runs on Node.js. Without it, your computer can't run the commands below.

Step 2. Install an AI coding tool. Pick one: Cursor, Antigravity, OpenCode or Claude Code. Install it and sign in. _Why:_ This is the AI that will build your app. Frame-Relay teaches it to use your designs.

Step 3. Add the Frame-Relay plugin to Figma. Open Figma, search for "Frame-Relay" in Community plugins, and click Install. Until the listing is live, download the plugin zip from the latest GitHub release and follow these import steps. _Why:_ The plugin is how your designs leave Figma and reach your AI.

Part 2: In Figma

Step 4. Design your components. Build your buttons, inputs and cards as Figma components. Follow the simple naming rules in the naming contract, such as naming a variant "Size=Large" or "State=Hover", and use Figma variables for your colors and spacing. _Why:_ Clear names let Frame-Relay understand your design. Variables let it carry your exact colors and sizes into code.

Step 5. Check your file. Run the Frame-Relay plugin and open the Lint tab. Click Check file. Fix anything it lists, then check again. _Why:_ It's like spell-check for your design file. It catches problems before they turn into messy code.

Step 6. Export your kit. Open the Export tab and click Export kit. A zip file downloads. _Why:_ This zip is your whole design system packed into one box: colors, sizes, components and a picture of each one.

Part 3: Set up your app

Step 7. Create a project folder. Make a new folder on your computer, for example my-app. Use dashes, not spaces, in the name. Open the folder in your AI tool. _Why:_ Your app's code needs a home.

Step 8. Ask your AI to set up the app. Paste this into your AI tool:

Set up a Vite + React + TypeScript app in this folder, with Tailwind CSS v4 and the "@/" path alias pointing to src. Then run npm run build to confirm it works.

_Why:_ This creates an empty app that Frame-Relay knows how to work with.

Step 9. Set up Frame-Relay. Open the terminal in your AI tool and type:

npx @frame-relay/cli@beta init

Answer yes to its questions. _Why:_ It looks at your app and gets it ready to receive your designs.

Step 10. Bring in your designs. Type this, with a space at the end:

npx @frame-relay/cli@beta sync --from

Then drag your zip file from Step 6 into the terminal and press Enter. _Why:_ This opens the box. It turns your Figma components into real code, and writes rules telling your AI: "Always use these. Never invent your own."

Step 11. Run a health check.

npx @frame-relay/cli@beta doctor

You should see green ticks. _Why:_ It's like the warning lights in a car. It confirms everything is connected before you start building.

Part 4: Build with your AI

Step 12. Ask your AI to build a screen. Start a new chat in your AI tool and ask for what you need, for example: "Build a login page using my design kit." _Why:_ The AI now looks up your components before it builds, so it uses your buttons and inputs instead of making up its own.

Step 13. See your app. In the terminal, type:

npm run dev

Open the link it shows, usually http://localhost:5173, in your browser. _Why:_ You see the page the AI built and can compare it with your Figma design.

Step 14. Check the AI's work.

npx @frame-relay/cli@beta check

_Why:_ It catches any place the AI broke your rules, such as using a random color. If it finds something, ask the AI: "Fix the issues frame-relay check found."

Part 5: Extras

Step 15. Use live mode (optional). Ask your AI: "Start Frame-Relay live mode." It gives you a 6-digit code. In Figma, open the plugin's Live tab, type the code and click Connect. Now select anything in Figma and ask your AI: "Build something that matches my Figma selection." _Why:_ Your AI sees what you select in Figma right away, so you can build from a design you haven't exported yet.

Step 16. When you change your designs. Repeat Steps 5, 6 and 10: check, export and sync again. _Why:_ Your code stays matched to your latest Figma design. If you edited a generated component by hand, Frame-Relay keeps your edit and saves the new version beside it, so you never lose work.

Supported AI tools

sync writes the setup for each tool it finds, so the agent can start using the kit right away.

ToolWhat sync writes
Antigravity.agents/mcp_config.json and .agents/rules/frame-relay.md
OpenCodeopencode.json (merged, other keys kept) and AGENTS.md
Cursor.cursor/mcp.json and .cursor/rules/frame-relay.mdc
Claude Code.mcp.json and CLAUDE.md

Until the stable release, the package is published with the beta tag only. The simplest setup is to install it as a dev dependency once: npm install -D @frame-relay/cli@beta. Sync then writes a command that uses the local install, and the MCP server starts without changes.

Documentation

PageWhat it covers
Getting startedFirst-time walkthrough from an empty folder to an AI-built page.
ConceptsThe kit, sync, rules files, MCP, check, doctor and live mode.
CLI commandsEvery command, its flags and the files it writes.
Figma pluginThe Lint, Export and Live tabs, plus warnings and the icon rule.
Live modePairing, the Live tab, live tools, security and troubleshooting.
MCP serverThe 10 tools your agent calls, resources and client setup.
Kit formatFolder layout and every field in the kit JSON files.
ConverterHow Figma data maps to kit specs, including every warning code.
Figma naming contractThe naming rules that keep exports clean.
JSON schemasStable URLs for the manifest, tokens and component schemas.

The same docs are published at https://joseph-brendan.github.io/Frame-Relay/.

Contributing

Setup, commands, branch naming and release notes are in CONTRIBUTING.md. Security reports: SECURITY.md. What we collect: nothing, see PRIVACY.md.

License

MIT, copyright (c) 2026 Joseph Brendan.

问 PopNx