OpenAI Agent

About

This .NET sample demonstrates the creation of an AI agent that leverages Azure OpenAI models and communicates with a remote MCP server, showcasing a complete client-server architecture. The client application, built with .NET Blazor, interacts with a TypeScript-based MCP server, both deployed on Azure Container Apps (ACA). Security is enforced through ACA's built-in authentication for the client, while the server is exclusively accessible by the client, ensuring a secure and isolated environment for AI agent interactions.

Key Features

• MCP client app is written in .NET Blazor.
• MCP server app is only accessible from the MCP client app.
• Connects to a to-do MCP server written in TypeScript.
• MCP client app is secured by the built-in auth of ACA.
• Both MCP client and server apps are running on Azure Container Apps (ACA).
• 11 GitHub stars

Use Cases

• Creating a to-do list agent that uses natural language prompts
• Securing communication between AI agents and MCP servers
• Deploying AI agents on Azure Container Apps