Qanat Goose

Qanat Goose MCP Extension is an agent‑based Goose MCP‑UI extension that wraps the Square MCP Server, integrating sandbox catalog and orders with voice and gesture agents for an interactive seller dashboard within Goose Desktop. This setup lets you manage your Square catalog and orders directly from the Goose UI while leveraging hands‑free and gesture controls to speed up common tasks.

How to use

You run the MCP server and connect Goose Desktop to the Qanat Goose MCP Extension. Once connected, you can control the Catalog and Orders dashboards using voice commands and gestures, or interact through the Goose Desktop UI. Use the extension to refresh catalogs, view recent orders, and toggle item statuses with simple gestures.

How to install

Prerequisites you need installed on your machine before you begin:

• Node.js and npm are required to run the MCP server in development mode.
• Python is required to run the Python variant of the MCP server.
• Git is helpful for cloning repositories.
• A Goose Desktop installation is required to host the MCP‑UI extension.

Step by step commands to set up and run the MCP server and extension:

# 1. Prepare environment (choose one of the run methods below)
# Copy environment template
cp env.example .env

# Edit .env with your keys
# SQUARE_API_KEY=your_key_here
# ELEVENLABS_API_KEY=your_key_here

# 2. Install dependencies (Python backend uses pip; frontend uses npm)
pip install -r requirements.txt

# 3. Run the MCP Server (choose one)
npm run dev

# Or run the Python MCP server variant
python -m backend.mcp_servers.qanat_server

Additional configuration and setup notes

Connect Goose Desktop to the Qanat Goose MCP Extension by adding it as an MCP‑UI extension and pointing to the local server endpoint. The extension exposes the MCP UI on the local host, ready to integrate with Goose Desktop.

When running via the MCP server, you can share the local endpoint with Goose Desktop as http://localhost:3001 so the extension can load within the Goose UI environment.

Configuration and environment variables

Required environment variables to enable APIs and features are listed below. Copy these into your .env file and ensure they are available to the running MCP server.

SQUARE_API_KEY=your_key_here
SQUARE_ENVIRONMENT=sandbox
ELEVENLABS_API_KEY=your_key_here
MEDIAPIPE_CONFIDENCE_THRESHOLD=0.65
`

Usage and commands

Voice commands and gesture controls are available to interact with the Catalog and Orders views. Use voice to refresh the catalog or show orders, and use gestures to toggle item status or switch views.

Voice commands include refreshing the catalog and displaying recent orders. Gesture controls include a thumb‑up to toggle item status, a point to select a row, an open palm to refresh, and a peace sign to switch between catalog and orders views.

Testing and validation

Run tests to ensure the integration works as expected. Use the test suite to verify Square API integration, voice agent, and gesture agent behavior.

Available tools

MCP-UI Server

Backend server that provides the MCP protocol surface and extension integration for Goose Desktop.

Square Integration

Module that communicates with Square API to fetch catalog data and orders.

Voice Agent

Speech-to-text and command interpretation using ElevenLabs services.

Gesture Agent

Gesture recognition using MediaPipe to trigger UI actions.