AI-Assisted Workflows¶

Vexy Lines exposes its MCP tools to AI assistants through the vexy-lines-mcp bridge installed by vexy-lines-cli.

Install¶

pip install vexy-lines-cli
vexy-lines-mcp --help

Launch Vexy Lines before using the bridge. The bridge connects the assistant's stdio MCP protocol to the app's local TCP server on localhost:47384.

Claude Desktop Configuration¶

Add this to ~/Library/Application Support/Claude/claude_desktop_config.json on macOS, or the matching Claude config path on Windows:

{
  "mcpServers": {
    "vexy-lines": {
      "command": "vexy-lines-mcp"
    }
  }
}

Restart Claude Desktop after editing the file.

Cursor Configuration¶

Add the same server under mcp.servers in Cursor's MCP configuration:

{
  "mcp": {
    "servers": {
      "vexy-lines": {
        "command": "vexy-lines-mcp"
      }
    }
  }
}

What The Assistant Can Do¶

The bridge exposes the app's MCP tools: document open/save/export, layer tree inspection, group/layer/fill creation, fill parameter edits, image-filter edits, source image changes, masks, transforms, render control, undo/redo, and selection.

It can only operate on the running app and the files the app can access. Treat AI-driven edits like any other automation: save copies, review output, and keep backups for important artwork.

Previous: Video to Vector Video | Next: Performance and Troubleshooting