Skip to content

first-agent.md

Latest commit

 

History

History
170 lines (123 loc) · 3.96 KB

first-agent.md

File metadata and controls

170 lines (123 loc) · 3.96 KB

2️⃣ Your First Agent

Let's build a simple AI agent using the Microsoft Agent Framework.

🎯 What You'll Build

A conversational agent that:

  • Connects to Azure OpenAI
  • Uses secure authentication (no API keys!)
  • Responds to user queries

📝 Create the Agent

Create a new file called my_agent.py:

"""
My First Azure AI Agent
"""
import asyncio
import os
from agent_framework.azure import AzureOpenAIResponsesClient
from azure.identity import DefaultAzureCredential

async def main():
    # Use managed identity / Azure CLI credentials (no API keys!)
    credential = DefaultAzureCredential()
    
    # Create the agent
    agent = AzureOpenAIResponsesClient(
        endpoint=os.environ.get("AZURE_OPENAI_ENDPOINT"),
        credential=credential,
    ).as_agent(
        name="MyFirstAgent",
        instructions="""You are a helpful AI assistant. 
        You provide clear, concise answers.
        You're friendly and professional."""
    )
    
    # Run the agent
    print("🤖 Agent is ready! Type 'quit' to exit.\n")
    
    while True:
        user_input = input("You: ").strip()
        
        if user_input.lower() in ["quit", "exit", "q"]:
            print("Goodbye! 👋")
            break
        
        if not user_input:
            continue
        
        response = await agent.run(user_input)
        print(f"\n🤖 Agent: {response}\n")

if __name__ == "__main__":
    asyncio.run(main())

🔧 Configure Environment

Set your Azure OpenAI endpoint:

# Set environment variable
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com"

# Login to Azure (for DefaultAzureCredential)
az login

▶️ Run the Agent

python my_agent.py

Expected output:

🤖 Agent is ready! Type 'quit' to exit.

You: Hello! What can you do?

🤖 Agent: Hello! I'm an AI assistant that can help you with a variety of tasks...

You: quit
Goodbye! 👋

🎉 Congratulations!

You just built your first AI agent! Let's understand what happened:

Component What It Does
DefaultAzureCredential Authenticates without API keys (uses Azure CLI, managed identity, etc.)
AzureOpenAIResponsesClient Connects to Azure OpenAI service
.as_agent() Creates an agent with name and instructions
agent.run() Sends a message and gets a response

🔧 Add Tools (Optional)

Agents become powerful when they can use tools. Here's a quick example:

from agent_framework.tools import function_tool

@function_tool
def get_current_time() -> str:
    """Get the current date and time."""
    from datetime import datetime
    return datetime.now().strftime("%Y-%m-%d %H:%M:%S")

@function_tool
def calculate(expression: str) -> str:
    """Calculate a mathematical expression."""
    try:
        # Safe evaluation of math expressions
        allowed = set("0123456789+-*/.() ")
        if all(c in allowed for c in expression):
            return str(eval(expression))
        return "Invalid expression"
    except:
        return "Error calculating"

# Add tools to your agent
agent = AzureOpenAIResponsesClient(
    endpoint=os.environ.get("AZURE_OPENAI_ENDPOINT"),
    credential=credential,
).as_agent(
    name="ToolAgent",
    instructions="You are a helpful assistant with access to tools.",
    tools=[get_current_time, calculate]  # 👈 Add tools here
)

Now your agent can:

  • Tell the time: "What time is it?"
  • Do math: "What's 15% of 250?"

🐛 Troubleshooting

Problem Solution
AZURE_OPENAI_ENDPOINT not set Run export AZURE_OPENAI_ENDPOINT="https://..."
AuthenticationError Run az login to refresh credentials
ResourceNotFound Check your endpoint URL is correct
RateLimitError Wait a moment and try again

➡️ Next Step

Your agent works locally! Let's deploy it to Azure for production use.

Continue to: Deploy to Azure →