mcp

Enneagora is a dual-implementation MCP server for e-commerce customer support. It exposes 14 built-in MCP tools across four categories, including order management (e.g., get_order_status, cancel_order, process_return, track_package), information and support (e.g., get_support_info, get_return_policy, get_shipping_info, get_contact_information), product guidance (e.g., get_size_guide, get_warranty_information, get_product_care_info), and account/payment helpers (e.g., get_payment_information, get_account_help, get_loyalty_program_info). The server is designed to be platform-agnostic, using a strategy layer to integrate with different e-commerce platforms and clients, such as Claude Desktop via STDIO or a Gradio web interface. For deployment, you can run the Gradio server locally or deploy the FastMCP STDIO variant for Claude Desktop integration, and you can also deploy the hosted Gradio version on Hugging Face Spaces. When connected to MCP clients, you can issue natural language queries like “Check the status of order ORD-1001-S” or “What’s the return policy for electronics?” and Enneagora will route the request to the appropriate tool with mock data support for testing.

Prerequisites:

• Python 3.10 or higher
• pip
• Virtual environment tooling (recommended)
• Optional: Hugging Face account for Spaces deployment

Installation steps:

1. Clone the repository

git clone https://github.com/slavpilus/mcp.git
cd mcp

2. Create and activate a virtual environment

python -m venv venv
# macOS/Linux
source venv/bin/activate
# Windows
venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. (Optional) Install development dependencies for testing

pip install -r requirements-dev.txt

5. Run the Gradio MCP server locally (development)

python main.py

6. Run the STDIO FastMCP variant for Claude Desktop integration (if desired)

python main_stdio.py

7. If you plan to deploy on Hugging Face Spaces, follow the hosting steps described in the README and ensure environment variables are configured as needed.

Tips and notes:

• The repository provides two MCP server implementations: Gradio (main.py) and FastMCP STDIO (main_stdio.py). Use the Gradio server for web UI and testing, or the STDIO variant when integrating with Claude Desktop.
• Environment variables can be defined per server entry if you need to configure API keys or platform-specific settings; the sample mcp_config includes empty env mappings that you can populate.
• Mock data supports pattern-based order scenarios (e.g., ORD-1001-S for shipped with tracking) for realistic testing without real transactions.
• If you encounter port or interface issues, verify that you are activating the correct Python environment and that no other service is occupying port 7860 (default for Gradio).
• For production deployment, consider the Hugging Face Spaces workflow described in the README and set up CI/CD secrets accordingly.