1. Conduid
  2. Other
  3. Whiteboard MCP
MCP server · Other

Whiteboard MCP

Unclaimed last commit 6 months ago other
35Low

Scored 4 days ago · breakdown

About Whiteboard MCP

Whiteboard MCP is an MCP server published by Whiteboard-MCP in the Other category.

The repository has 12 stars and 2 forks, with the last commit 6 months ago. Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.

Install

From source
git clone https://github.com/Whiteboard-MCP/Whiteboard-MCP

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about Whiteboard MCP

Powered by Claude · Grounded in docs

I know everything about Whiteboard MCP. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

Static checks derived from the repository 2 days ago. Heuristics on the source, not an audit.

  • ✓README presentThe repository ships documentation.
  • !License declaredNo license file. Legally you have no right to use it in a product.
  • !Tests presentNo tests found. Behaviour changes may go unnoticed between versions.
  • !Dependencies pinnedNo lockfile. Installs can pull different code over time.

2 checks have not been run yet.

README

Whiteboard MCP

Give your AI the power to draw. Architecture diagrams, flowcharts, wireframes, sketches — generated from a single chat message.

whiteboard-mcp.com — Get started in 30 seconds

npm version

Whiteboard MCP is a Model Context Protocol server that gives AI agents like Claude a real canvas to draw on. Your AI describes what to draw, and you watch it appear in real-time.


What is Whiteboard MCP?

Whiteboard MCP connects your AI coding agent to an interactive drawing canvas. Instead of describing diagrams in text or generating static images, the AI draws directly on a whiteboard that you can see, interact with, and export.

One prompt. One diagram. Zero manual drawing.

> Look at this codebase and draw me an architecture diagram

Claude: I'll analyze the codebase and create a diagram.
  ▸ open_canvas()
  ▸ create_flowchart(nodes: [...], edges: [...])
  ▸ take_screenshot()

Here's your architecture diagram showing the API Gateway
routing to Auth, Core, and Worker services.

The canvas opens in your browser. Shapes appear in real-time as the AI draws. You can pan, zoom, select, move, resize — it's a full whiteboard.


Quick Start

The easiest way

Just paste this into your AI assistant (Claude Code, Cursor, etc.):

Install the whiteboard-mcp MCP server for me. You can find install instructions at https://whiteboard-mcp.com — add it to my MCP config so I can use it. Then open the canvas and draw me a quick test diagram.

Your AI should handle the rest. Restart your AI assistant afterwards.

Manual install

For full install instructions with editor-specific configs, visit whiteboard-mcp.com.

Or add Whiteboard MCP to your editor's MCP configuration directly:

Project-level — add to .mcp.json in your project root:

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

Global — add to ~/.claude/settings.json to make it available in all projects:

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

Restart Claude Code. Then just ask:

"Open the whiteboard and draw me a flowchart of the login process"

Edit your Claude Desktop config:

  • Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "whiteboard": {
      "command": "npx",
      "args": ["-y", "whiteboard-mcp"]
    }
  }
}

Add to .cursor/mcp.json in your project root:

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

Add to .vscode/mcp.json in your project root:

{
  "servers": {
    "whiteboard": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "whiteboard-mcp"]
    }
  }
}

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

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

Add to your Zed settings.json under context_servers:

{
  "context_servers": {
    "whiteboard": {
      "command": {
        "path": "npx",
        "args": ["-y", "whiteboard-mcp"]
      }
    }
  }
}

Add to opencode.json in your project root, or ~/.config/opencode/opencode.json for global:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "whiteboard-mcp": {
      "type": "local",
      "command": ["npx", "-y", "whiteboard-mcp"],
      "enabled": true
    }
  }
}

Note: OpenCode uses "mcp" (not "mcpServers") and command is an array.

Add to .amp/settings.json in your project root, or ~/.config/amp/settings.json for global:

{
  "amp.mcpServers": {
    "whiteboard-mcp": {
      "command": "npx",
      "args": ["-y", "whiteboard-mcp"]
    }
  }
}

Add to .roo/mcp.json in your project root:

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

Open Cline settings > MCP Servers > Edit Global MCP, then add:

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

Add to .amazonq/mcp.json in your project root, or ~/.aws/amazonq/mcp.json for global:

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

Run the server directly:

npx -y whiteboard-mcp

Or configure your MCP client to launch it with command: "npx" and args: ["-y", "whiteboard-mcp"] using stdio transport.

Try it

After installing, tell your AI:

  • "Open the whiteboard and draw me an architecture diagram of this project"
  • "Draw a flowchart showing the user registration process"
  • "Sketch a dog sniffing some flowers in the sunshine"
  • "Create a diagram with a web client, API gateway, auth service, and database"

Features

30+ Drawing Tools

Category Tools
Shapes draw_rectangle draw_ellipse draw_diamond
Lines draw_line draw_arrow draw_freehand
Content add_text add_image
Diagrams add_labeled_box connect_shapes create_flowchart
Manipulation move_element resize_element update_element delete_elements duplicate_elements
Layer bring_to_front send_to_back
Canvas open_canvas clear_canvas set_viewport set_canvas_background toggle_grid
Export take_screenshot save_png save_svg save_json export_json export_svg
History undo redo
Query get_canvas_info get_elements find_elements

One-Call Diagrams

The create_flowchart tool generates complete auto-laid-out diagrams from a simple graph definition:

create_flowchart({
  nodes: [
    { id: "client", label: "Web Client" },
    { id: "api", label: "API Gateway" },
    { id: "db", label: "PostgreSQL", type: "ellipse" }
  ],
  edges: [
    { from: "client", to: "api", label: "HTTPS" },
    { from: "api", to: "db", label: "SQL" }
  ],
  layout: "TB"
})

One tool call. Complete diagram. Auto-positioned nodes, labeled arrows, proper spacing.

Smart Arrows

connect_shapes creates arrows that bind to shapes. When you move a shape, the arrow follows. The AI builds diagrams with connected components, not disconnected shapes.

Interactive Canvas

The canvas isn't just for the AI — you can use it too:

  • Draw — Rectangle, ellipse, diamond, line, arrow, freehand, text tools
  • Select & Move — Click to select, drag to move
  • Resize — Drag corner handles to resize any element
  • Pan & Zoom — Scroll to pan, Ctrl+scroll to zoom, hand tool for drag-pan
  • Snap — Snap to grid and snap to shapes for precise positioning
  • Keyboard shortcuts — V (select), R (rectangle), O (ellipse), D (diamond), L (line), A (arrow), P (pen), T (text), H (hand), Delete, Ctrl+Z/Y

Export

  • PNG — Raster export for sharing and documentation
  • SVG — Vector export for high-quality printing and editing
  • JSON — Save and load your canvas state

Preferences

  • Canvas background color (including dark mode)
  • Grid visibility toggle
  • Snap to grid
  • Snap to shapes

How It Works

Whiteboard MCP runs as a local MCP server on your machine. When your AI agent calls open_canvas, it starts a lightweight web server and opens the canvas in your browser. All drawing happens through MCP tool calls — the AI sends commands, the canvas renders them in real-time via WebSocket.

AI Agent (Claude, Copilot, etc.)
    │
    │ MCP protocol (stdio)
    ▼
Whiteboard MCP Server
    │
    ├── HTTP server (serves canvas UI)
    ├── WebSocket (real-time state sync)
    │
    ▼
Browser Canvas (your whiteboard)

Everything runs locally. No cloud services, no data leaving your machine, no accounts to create.


Requirements

  • Node.js 18+ (for npx)
  • Any modern browser (Chrome, Firefox, Edge, Safari)
  • An MCP-compatible AI agent (Claude Code, Claude Desktop, Cursor, VS Code, Windsurf, Zed, etc.)

Pricing

Whiteboard MCP includes a 14-day free trial — no credit card required. After the trial, purchase a license for a one-time payment of $29.

Buy a license

Activating Your License

After purchasing, add your license key to your MCP config:

{
  "mcpServers": {
    "whiteboard": {
      "command": "npx",
      "args": ["-y", "whiteboard-mcp"],
      "env": {
        "WHITEBOARD_MCP_LICENSE_KEY": "YOUR-KEY-HERE"
      }
    }
  }
}

Restart your editor. The trial banner changes to PRO.


Support


FAQ

Yes. Claude Code is the primary target. Add the MCP config and Claude can draw diagrams, flowcharts, and sketches directly.

Yes. Whiteboard MCP works with any editor or AI agent that supports the Model Context Protocol (MCP). See the install instructions above for editor-specific configuration.

No. npx whiteboard-mcp downloads and runs everything automatically. No global installs, no build steps, no dependencies to manage.

No. Everything runs locally on your machine. The only network call is license key validation (a single POST to verify your key). Your drawings, code, and data never leave your computer.

The canvas has a full interactive UI — you can draw shapes, move elements, and export manually. But the real power is having an AI drive it. That's what Whiteboard MCP is built for.

After 14 days, the tools still work but you'll see periodic reminders to purchase a license. Buy once for $29 and own it forever — lifetime updates included.

Yes. If Whiteboard MCP doesn't work for you, email whiteboard-mcp@fastmail.com within 30 days for a full refund.


Links


Built for developers who think visually.

README fetched from the repository 2 days ago. The original is authoritative.

Questions

About Whiteboard MCP

How do I install Whiteboard MCP?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is Whiteboard MCP safe to use with an AI agent?

Its trust score is 35 out of 100 (low). It passes 1 of the 4 static security checks run so far; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Whiteboard MCP still maintained?

The last commit was 6 months ago and 2 open issues. That's long enough that you should check whether the maintainer is responding to issues before depending on it.