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)
What fits, with live prices: specs, fit verdicts with rule ids and stock for 3D printing.
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。
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 …”。
适用于网页版、桌面版和手机版 Claude:
https://compatsy.com/3d-printing/mcp在终端中运行这条命令
claude mcp add --transport http 3d-printing https://compatsy.com/3d-printing/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"3d-printing": {
"url": "https://compatsy.com/3d-printing/mcp"
}
}
}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": {
"3d-printing": {
"type": "http",
"url": "https://compatsy.com/3d-printing/mcp"
}
}
}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。
ChatGPT 只能连接互联网上的服务器:
https://compatsy.com/3d-printing/mcp菜单名称可能因套餐和版本不同而略有差异。
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。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
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6日
注册名称已通过域名 compatsy.com 验证(注册中心会核实域名所有权)
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
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
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。
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。
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(0.1.0)于7天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
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。
compatsy answers "does this part fit that machine, and where is it in stock?" Each category has its own MCP endpoint:
| Category | Covers | Endpoint |
|---|---|---|
| 3D printing | printers, nozzles, hotends, build plates, filament, enclosures | https://compatsy.com/3d-printing/mcp |
| Home espresso | machines, tampers, puck screens, distribution tools, dosing funnels, baskets | https://compatsy.com/home-espresso/mcp |
Specs carry their source and confidence, fit verdicts cite the rule that decided them, and prices come with the time they were last verified.
This repo is the quickstart for the hosted MCP server. Nothing runs locally except an optional stdio bridge.
Remote (preferred). Any client that supports Streamable HTTP:
{ "mcpServers": { "compatsy-3d-printing": { "url": "https://compatsy.com/3d-printing/mcp" } } }
stdio via npx. For clients that only launch local servers:
{ "mcpServers": { "compatsy-espresso": { "command": "npx", "args": ["-y", "github:elliot-rosen/compatsy-mcp"], "env": { "COMPATSY_MCP_URL": "https://compatsy.com/home-espresso/mcp" } } } }
Claude Code: claude mcp add --transport http compatsy-3d-printing https://compatsy.com/3d-printing/mcp
No account or key. Calls are rate limited per session. The bridge defaults to https://compatsy.com/mcp (3D printing); set COMPATSY_MCP_URL to another endpoint from the table.
| Tool | What it does | Read-only |
|---|---|---|
search_products | Find products by name or category plus spec filters | yes |
get_product | Full specs with source and confidence, every offer with verified time | yes |
compare | Side-by-side spec table for 2 to 10 products | yes |
check_fit | Verdict for a part and a machine, with the rule id and version, or the missing fields | yes |
check_stock | Price and availability per merchant, re-verified when over an hour old | yes |
report_issue | Flag a wrong spec, price or verdict. Reports never edit facts | no |
{"name": "search_products", "arguments": {"query": "hardened nozzle 0.4", "category": "nozzle", "limit": 5}}
{"name": "get_product", "arguments": {"product_id": "prusa-mk4s"}} {"name": "compare", "arguments": {"product_ids": ["bambu-lab-p1s", "prusa-core-one"]}} {"name": "check_fit", "arguments": {"subject_id": " ", "target_id": "prusa-mk4s"}} {"name": "check_stock", "arguments": {"product_id": "prusa-mk4s"}} {"name": "report_issue", "arguments": {"entity_type": "product", "entity_id": "prusa-mk4s", "field": "nozzle_thread", "claim": "MK4S uses the Nextruder thread"}}
Example prompts that use them well:
fits, fits_with_adapter, does_not_fit or unknown. Unknown is an answer: it lists the spec fields that are missing rather than guessing. Every rule needs two independent sources, one from the manufacturer, before it can decide anything.
rules/cursor.mdc and rules/windsurf.md tell Cursor or Windsurf agents when to call compatsy instead of guessing fitment from memory.