Figma Context MCP
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
作者提供的说明(英文原文)
A local MCP server with full Figma REST API coverage.
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。
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。
选择你使用的 AI 应用,然后复制设置。
FIGMA_API_KEY标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
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": {
"figma": {
"command": "npx",
"args": [
"-y",
"figma-mcp-server"
],
"env": {
"FIGMA_API_KEY": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add figma --env FIGMA_API_KEY=your-value -- npx -y figma-mcp-server运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"figma": {
"command": "npx",
"args": [
"-y",
"figma-mcp-server"
],
"env": {
"FIGMA_API_KEY": "your-value"
}
}
}
}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。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"figma": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"figma-mcp-server"
],
"env": {
"FIGMA_API_KEY": "your-value"
}
}
}
}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。
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
注册名称已通过 GitHub 账户 planetabhi 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
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。
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。
你需要填写:FIGMA_API_KEY
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)。
GitHub 仓库最后更新于–,最新版本(3.1.3)于6天前进入注册中心。
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。
同一分类中有:Figma Context MCP, Claude Real Video, Figma MCP Server
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ 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
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
Let any LLM watch a video locally — and search everything it has ever watched.
The Figma MCP server brings Figma design context directly into your AI workflow.
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。
A single call leaves your editor and returns with a design laid open. This is a local server that carries the whole of the Figma REST API, every endpoint within reach and none withheld. It speaks with Claude Desktop, Cursor, VS Code, the Gemini CLI, and any client that speaks MCP. You will find it on npm and in the official MCP Registry, waiting to be called.
Every endpoint of the Figma REST API stands here as a single self-describing tool, and each one says plainly what it needs. At startup the server walks the tools/figma/ directory and gathers them, so to teach it a new endpoint you lay down one new file and nothing more. Before any call goes out its required parameters are weighed, and the unfinished request is turned back at the threshold. It speaks over stdio, quietly, one message at a time.
This server is grown from Bun, and it asks for version 1.2 or newer. Being Bun-native, it answers to bunx and not to npx, for Node cannot carry it.
One key opens the whole door. Create a personal access token in Figma under Settings > Security > Personal access tokens > Generate new token, then give it to the server as FIGMA_API_KEY, set within your client config's env block as shown below.
Many clients, and one server beneath them all. Point command at bunx, or at its absolute path from which bunx when the client cannot find it on your PATH.
{
"mcpServers": { "figma": { "command": "bunx", "args": ["figma-mcp-server"], "env": { "FIGMA_API_KEY": "your_figma_api_key_here" } } } }
Each client keeps its config in its own place.
claude_desktop_config.json)~/.cursor/mcp.json, or .cursor/mcp.json per project~/.gemini/settings.jsonOnce the config is written, restart the client so the change takes hold.
Use .vscode/mcp.json for a workspace, or your user mcp.json, each holding a servers key. Rather than carve the token into the file, name an input and let VS Code ask for it, so the secret is spoken once and then kept safe.
{
"inputs": [ { "type": "promptString", "id": "figma-api-key", "description": "Figma API Key", "password": true } ], "servers": { "figma": { "command": "bunx", "args": ["figma-mcp-server"], "env": { "FIGMA_API_KEY": "${input:figma-api-key}" } } } }
${input:figma-api-key}, VS Code prompts you for the key the first time the server starts, then stores it in your OS secret storage.The whole of the Figma REST API is here, every endpoint that still stands, fifty-five tools in all, gathered across these domains.
For the full and ever-current roll of tool names and their parameters, look to your MCP client's own list.
Most tools open to any plan's personal access token. A few ask for more.
list_file_variables, get_published_variables, modify_variables): Enterprise organization.get_library_*): Organization or Enterprise plan.duplicate_file, create_team, create_folder, invite_team_members): Enterprise plan via a private OAuth app.Figma's API does not stand still, and neither does this server. Twice each month, on the first day and the fifteenth, a workflow wakes of its own accord, reaches out for the newest Figma specification, and sets it beside the one already held, marking every endpoint that has arrived, changed, or departed. It runs the type checks and the tests before it trusts the change, and when all of it holds it opens a pull request for a human to read and welcome in. You may also call it by hand whenever the need arises. The work lives in .github/workflows/sync-spec.yml, and the spec it tends is kept in spec/figma-openapi.json.
MIT License · By @planetabhi ⋛⋋( ⊙◊⊙)⋌⋚