Design perfboards with your AI assistant

Claude Code

Run this in a terminal:

claude mcp add --transport http perfboard https://mcp.perfboarddesigner.com/mcp

Add --scope user to have it in every project, not just the current folder. Check it with /mcp inside Claude Code.

Claude.ai & Claude Desktop

1. Open Connectors in Claude's settings and choose Add custom connector.
2. Name it Perfboard Designer and paste the address:
   https://mcp.perfboarddesigner.com/mcp
3. Click Add. There's no sign-in, so leave the advanced (OAuth) settings empty.
4. In a chat, open the + menu, then Connectors, and turn Perfboard Designer on.

A connector added once works on claude.ai, in Claude Desktop and in the mobile apps. On Team and Enterprise plans an owner adds it for the organization first. Custom connectors depend on your Claude plan.

ChatGPT

1. In Settings → Apps & Connectors → Advanced settings, turn on Developer mode.
2. Back in Apps & Connectors, choose Create, name it Perfboard Designer and paste the address:
   https://mcp.perfboarddesigner.com/mcp
3. Set authentication to No authentication and create it.
4. In a chat, pick it from the + menu under Developer mode.

Developer mode is on paid ChatGPT plans, and OpenAI moves these menus from time to time.

Cursor

Add it to ~/.cursor/mcp.json (every project) or .cursor/mcp.json (one project):

{
  "mcpServers": {
    "perfboard": {
      "url": "https://mcp.perfboarddesigner.com/mcp"
    }
  }
}

VS Code (GitHub Copilot)

Add it to .vscode/mcp.json in your project, or run MCP: Add Server from the Command Palette:

{
  "servers": {
    "perfboard": {
      "type": "http",
      "url": "https://mcp.perfboarddesigner.com/mcp"
    }
  }
}

Then use it from Copilot Chat in agent mode.

Windsurf

Add it to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "perfboard": {
      "serverUrl": "https://mcp.perfboarddesigner.com/mcp"
    }
  }
}

Gemini CLI

gemini mcp add --transport http perfboard https://mcp.perfboarddesigner.com/mcp

Or add it to ~/.gemini/settings.json by hand. Note the key is httpUrl, not url:

{
  "mcpServers": {
    "perfboard": {
      "httpUrl": "https://mcp.perfboarddesigner.com/mcp"
    }
  }
}

Codex CLI

codex mcp add perfboard --url https://mcp.perfboarddesigner.com/mcp

Or add it to ~/.codex/config.toml:

[mcp_servers.perfboard]
url = "https://mcp.perfboarddesigner.com/mcp"

Other clients

Any client that supports remote MCP servers over Streamable HTTP works: give it the address https://mcp.perfboarddesigner.com/mcp, with no authentication.

For a client that can only start local (stdio) servers, bridge to it with mcp-remote, which needs Node.js:

{
  "mcpServers": {
    "perfboard": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.perfboarddesigner.com/mcp"]
    }
  }
}

Run it yourself

The server is also on npm as perfboard-designer-mcp, so you can run it on your own computer over stdio instead of using the hosted address. You need Node.js 22 or newer; npx downloads it the first time. In Claude Code:

claude mcp add --scope user perfboard -- npx -y perfboard-designer-mcp

Claude Desktop (Settings → Developer → Edit Config, then restart it), Cursor, Windsurf and other clients that start a command take the same in their JSON config:

{
  "mcpServers": {
    "perfboard": {
      "command": "npx",
      "args": ["-y", "perfboard-designer-mcp"]
    }
  }
}

Your boards stay on your computer. The server contacts Perfboard Designer's servers only when you ask for a short link, and the picture link it returns points at the hosted server. Settings go in the config's "env", next to "args":

• PERFBOARD_EDITOR_URL: where board links open, e.g. http://localhost:3000/app/ for your own copy of the designer.
• PERFBOARD_PROJECTS_URL and PERFBOARD_IMAGE_URL: set to off for no short links or no picture links.

"env": {"PERFBOARD_PROJECTS_URL": "off", "PERFBOARD_IMAGE_URL": "off"}

To run it from the source instead, clone the repo, build it (cd mcp && npm install && npm run build) and use node /path/to/perfboard-designer/mcp/dist/server.js as the command. The server's README also covers hosting your own copy on Cloudflare Workers.