Skip to content

MAF-HostedAgent-01-TimeZone

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

MAF-HostedAgent-01-TimeZone

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
Properties
Properties
 
 
Dockerfile
Dockerfile
 
 
MAF-HostedAgent-01-TimeZone.csproj
MAF-HostedAgent-01-TimeZone.csproj
 
 
Program.cs
Program.cs
 
 
README.md
README.md
 
 
agent.yaml
agent.yaml
 
 

README.md

MAF Hosted Agent 01 — TimeZone Agent

A .NET console agent built with the Microsoft Agent Framework (MAF) v1.0 that answers time and date questions using a local tool function. This sample demonstrates how to create a containerized agent suitable for deployment to Azure Foundry Agent Service as a Hosted Agent.

What This Sample Demonstrates

  • Creating an AI agent with MAF v1.0 (AsAIAgent(), RunAsync(), AgentResponse)
  • Registering a tool function (GetCurrentTime) that the agent can call
  • Using TimeZoneInfo to convert between timezones
  • Supporting both Azure OpenAI and Ollama as AI providers via environment variables
  • Packaging the agent as a Docker container for Hosted Agent deployment
  • Using the Foundry agent manifest (agent.yaml)

Prerequisites

Requirement Details
.NET 10 SDK Download
Docker Required for container build and Foundry deployment
Azure OpenAI or Ollama At least one AI provider configured
Azure subscription Required only for Azure OpenAI and Foundry deployment

Project Structure

MAF-HostedAgent-01-TimeZone/
├── Program.cs                    # Agent logic, tool function, chat loop
├── Dockerfile                    # Multi-stage Docker build
├── agent.yaml                    # Foundry agent manifest
├── MAF-HostedAgent-01-TimeZone.csproj
├── README.md
└── Properties/
    └── launchSettings.json       # Environment variable defaults

Getting Started

1. Configure an AI Provider

Option A — Azure OpenAI (set environment variables):

export AZURE_OPENAI_ENDPOINT="https://<your-resource>.openai.azure.com/"
export AZURE_OPENAI_MODEL="gpt-5-mini"
export AZURE_OPENAI_APIKEY="<your-api-key>"

Option B — Ollama (local, no Azure needed):

ollama pull phi4-mini
# Defaults: OLLAMA_ENDPOINT=http://localhost:11434/ OLLAMA_MODEL=phi4-mini

2. Build and Run Locally

cd samples/MAF/MAF-HostedAgent-01-TimeZone
dotnet build
dotnet run

You will see an interactive prompt:

TimeZone Agent — Ask me for the current time in any timezone!
Type 'exit' to quit.

You: What time is it in Tokyo?
Agent: The current time in Tokyo Standard Time is 2025-04-08 21:30:15 (Tokyo Standard Time).

3. Build the Docker Container

docker build -t timezone-agent:latest .

4. Run in Docker

docker run -it --rm \
  -e AZURE_OPENAI_ENDPOINT="https://<your-resource>.openai.azure.com/" \
  -e AZURE_OPENAI_MODEL="gpt-5-mini" \
  -e AZURE_OPENAI_APIKEY="<your-api-key>" \
  timezone-agent:latest

5. Deploy to Azure Foundry Agent Service

Follow the Hosted Agents deployment guide to push your container image and register the agent using the agent.yaml manifest.

Code Walkthrough

Chat Client Factory

The app checks environment variables to determine which AI provider to use. Azure OpenAI takes priority; if not configured, it falls back to Ollama:

IChatClient chatClient = CreateChatClient();

Agent with Tool Function

The agent is created using MAF's AsAIAgent() extension method. A GetCurrentTime tool function is registered via AIFunctionFactory.Create():

AIAgent timeAgent = chatClient.AsAIAgent(
    name: "TimeZoneAgent",
    instructions: "You are a helpful assistant that provides the current date and time...",
    tools: [AIFunctionFactory.Create(GetCurrentTime)]);

Tool Function

The GetCurrentTime function uses TimeZoneInfo.FindSystemTimeZoneById() to convert UTC to the requested timezone:

[Description("Gets the current date and time for the specified timezone.")]
static string GetCurrentTime(string timezoneName)
{
    var tz = TimeZoneInfo.FindSystemTimeZoneById(timezoneName);
    var now = TimeZoneInfo.ConvertTime(DateTimeOffset.UtcNow, tz);
    return $"Current time in {tz.DisplayName}: {now:yyyy-MM-dd HH:mm:ss}";
}

Interactive Chat Loop

The agent runs in a simple chat loop, processing each user message independently:

AgentResponse response = await timeAgent.RunAsync(input);
Console.WriteLine($"Agent: {response.Text}");

Related Resources