Canvashand
Start free trial

Connect your canvas in four clear steps.

Set up the account, MCP server, and Figma plugin once. After that, your AI client can work with the file you already have open.

Sign in to begin Usually takes five minutes

New to Terminal? On a Mac, press ⌘ Space, type Terminal, press Return. Paste a line, press Return again.

Three quick checks

  • Figma desktop is installed
  • You can install development plugins in Figma
  • Your AI client supports MCP

  1. Create your Canvashand account

    Use the same Google account in the dashboard and plugin so usage and billing stay attached to one workspace.

    Signed in
    Continue with Google

  2. Install the CLI and connect your AI client

    Install the Canvashand CLI, then run the setup command from the project where you want it available. It finds the AI clients you use (Claude Code, Codex, Cursor, VS Code, Gemini CLI, Kiro, Zed, Qwen Code, OpenCode, Antigravity) and installs the MCP connection and design workflow skill for each. For Windsurf or Cline, add --scope user.

    AI client connected
    curl -fsSL https://www.canvashand.com/install.sh | sh
    canvashand setup --all --scope project
    Manual downloads

  3. Install the Figma plugin

    The installer puts the plugin in ~/.canvashand/canvashand-plugin, and canvashand upgrade keeps it current. Import its manifest once in the Figma desktop app. If you imported Canvashand from a zip before, remove that plugin in Figma first. The plugin is not on Figma Community — it ships with your account.

    Plugin in Figma
    1. Figma desktop → Plugins → Development
    2. Import plugin from manifest… → pick ~/.canvashand/canvashand-plugin/manifest.json
    3. The folder is hidden, so type its path in the file picker. macOS: press ⌘ Cmd+Shift+G and paste the path. Windows: press Alt+D and paste %USERPROFILE%\.canvashand\canvashand-plugin. Linux: press Ctrl+L and paste the path.
    Or download canvashand-plugin.zip
    Importing the plugin on macOS. File names and designs are blurred.

  4. Open the bridge and test it

    Run Canvashand inside the file, keep the plugin window open, then ask your AI client to inspect the current selection.

    Bridge open

    “Use Canvashand to describe my selected frame.”

You are ready to design with your agent.

Select a frame, describe the change, and let Canvashand carry the document context between the live canvas and your AI client.

Open dashboard Manual downloads Claude Code guide Cursor guide