What is Azure Openai Agent Multi Mcp Starter

azure-openai-agent-multi-mcp-starter is a starter template designed for building Model Context Protocol (MCP) agents using Azure OpenAI. It serves as a foundation for developing advanced AI agents that can interact with external tools and services through the MCP protocol.

Use cases

Use cases include building intelligent chatbots, creating automated data processing agents, developing tools for customer support, and implementing AI solutions that require integration with multiple APIs and services.

How to use

To use azure-openai-agent-multi-mcp-starter, clone the repository, install the required packages using ‘pip install -r requirements.txt’, and configure your environment variables in the .env file with your Azure OpenAI API key, endpoint, and version.

Key features

Key features include the ability to utilize the Model Context Protocol (MCP) for enhanced interactions, integration with LangChain for seamless agent communication, and customizable MCP server options for tailored tool support.

Where to use

azure-openai-agent-multi-mcp-starter can be used in various fields such as AI development, automation, customer service, and any application requiring advanced AI interactions with external systems.

azure-multi-mcp-starter

This project is a starter template for building Model Context Protocol (MCP) agents using Azure OpenAI, not OpenAI. It demonstrates how to use the MCP protocol with Azure’s hosted models, providing a foundation for developing advanced AI agents that can interact with external tools and services.

What is MCP?

Model Context Protocol (MCP) is an open protocol that standardizes how language models interact with external tools, APIs, and environments. MCP enables models to call functions, access data, and perform actions beyond simple text generation, making them more useful and interactive in real-world applications.

LangChain MCP Adapter

The LangChain MCP adapter is a library that bridges LangChain agents with MCP servers. It allows you to use LangChain’s agent framework to communicate with any MCP-compliant server, enabling seamless integration of tool use, function calling, and advanced agent behaviors.

MCP Options

There are several MCP server implementations and options available:

• Python MCP Server: Run your own MCP server in Python, customizing tool support and logic.
• LangChain MCP Adapter: Use LangChain’s adapter to connect agents to MCP servers.

About This Template

This repository is designed as a quick-start template for developers building MCP agents with specifically Azure OpenAI, as other guides only provided template for OpenAI endpoints. It includes sample code for connecting to an MCP server, using the LangChain MCP adapter, and running an agent that can interact with external tools. You can create your own sample MCP Server with MCP tools for your AI to perform agentic actions. It is intended for Azure OpenAI users and requires adaptation to work with OpenAI’s public API.

Quickstart

To run this template follow the steps:

1. Clone repository
2. ```` pip install -r requirements.txt ```
3. Populate your .env file with the following: AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT,AZURE_OPENAI_VERSION
4. In terminal, run python frontend_api.py
5. This will start all the MCP servers and the LangChain MCP adapter.

Sample application

Upcoming features

Include more baseline MCP server integration with Azure AI Foundry, Azure MCP, Fabric MCP Server

For more information, see the official documentation for MCP, LangChain, and Azure OpenAI.