agentic-shell-python: Skill for Claude Code

Microsoft Agent Framework Integration

This skill helps you integrate the Microsoft Agent Framework into a Python FastAPI application for building AI agents powered by Azure OpenAI.

When to Use

• Adding AI agent capabilities to a Python backend
• Building conversational AI with Azure OpenAI
• Creating agents that can use tools and follow instructions
• Setting up the backend for AG-UI protocol support

Prerequisites

• Python 3.11 or higher
• An Azure OpenAI resource with a deployed model
• Azure CLI authenticated (az login)

Step-by-Step Instructions

1. Install Dependencies

Add these dependencies to your pyproject.toml:

[project]
dependencies = [
    "agent-framework-ag-ui>=1.0.0b251120",
    "fastapi>=0.115.0",
    "uvicorn>=0.32.0",
]

Then install with:

cd src/agentic-api
uv pip install -e .

2. Configure Environment Variables

Create a .env file or set these environment variables:

AZURE_OPENAI_ENDPOINT=https://<your-resource>.openai.azure.com/
AZURE_OPENAI_DEPLOYMENT_NAME=<your-deployment-name>

3. Update main.py

Replace or update your main.py with the agent framework setup:

import contextlib
import logging
import os

import fastapi
import telemetry

from agent_framework import ChatAgent
from agent_framework.azure import AzureOpenAIChatClient
from agent_framework_ag_ui import add_agent_framework_fastapi_endpoint
from azure.identity import AzureCliCredential

# Read required configuration
endpoint = os.environ.get("AZURE_OPENAI_ENDPOINT")
deployment_name = os.environ.get("AZURE_OPENAI_DEPLOYMENT_NAME")

if not endpoint:
    raise ValueError("AZURE_OPENAI_ENDPOINT environment variable is required")
if not deployment_name:
    raise ValueError("AZURE_OPENAI_DEPLOYMENT_NAME environment variable is required")

# Create the Azure OpenAI chat client with managed identity
chat_client = AzureOpenAIChatClient(
    credential=AzureCliCredential(),
    endpoint=endpoint,
    deployment_name=deployment_name,    
)

# Create the AI agent
agent = ChatAgent(
    name="AGUIAssistant",
    instructions="You are a helpful assistant.",
    chat_client=chat_client,
)

@contextlib.asynccontextmanager
async def lifespan(app: fastapi.FastAPI):
    telemetry.configure_opentelemetry()
    yield

app = fastapi.FastAPI(lifespan=lifespan)

# Register the AG-UI endpoint at the root
add_agent_framework_fastapi_endpoint(app, agent, "/")

logger = logging.getLogger(__name__)