Openai MCP Example

🚀 Azure Container Apps AI MCP Project Documentation

This is an application designed to interact with Azure containers, GitHub models, and the OpenAI API. The project offers the following features:

• Connect to different AI services via the MCP protocol.
• Support multiple authentication methods (including keyless authentication).
• Provide a complete development environment for easy integration and testing.
• Include detailed documentation and sample code.

🚀 Quick Start

Run the Project

1. Clone the repository:

git clone https://github.com/manekinekko/azure-container-apps-ai-mcp.git
cd azure-container-apps-ai-mcp

2. Start the Docker container:

docker compose up -d --build

3. Enter the MCP console:

docker exec -it mcp-host bash

📦 Installation Guide

Environment Variable Configuration

Set the corresponding environment variables according to the AI service provider you choose.

Azure OpenAI Configuration

Add the following content to the .env file:

AZURE_OPENAI_ENDPOINT="https://<ai-foundry-openai-project>.openai.azure.com"
MODEL="gpt-4.1"

Optional: If you don't want to use keyless authentication, you can set the API key:

AZURE_OPENAI_API_KEY=your_azure_openai_api_key

OpenAI Configuration

Add the following content to the .env file:

OPENAI_API_KEY=your_openai_api_key
MODEL="gpt-4.1"

GitHub Model Configuration

Add the following content to the .env file:

GITHUB_TOKEN=your_github_token
MODEL="openai/gpt-4.1"

💻 Usage Examples

Use DevContainer (Recommended)

You can directly start the development environment using GitHub Codespaces or the remote container feature of Visual Studio Code:

Run Docker Directly

1. Build and start the container:

docker compose up -d --build

2. Enter the MCP console:

docker exec -it mcp-host bash

📚 Documentation

Debugging and Logging

You can enable detailed log output by setting the DEBUG environment variable:

DEBUG=mcp:* npm start --prefix mcp-host

The debugging function is enabled by default, and you can view detailed runtime information in the console.

📄 License

This project is licensed under the MIT License. See the LICENSE file for the full license text.