agent-framework-azure-ai-py - Build Persistent Agents on Azure AI Foundry with Agent Framework SDK
Create and manage persistent agents on Azure AI Foundry using the Microsoft Agent Framework Python SDK with hosted tools, function tools, MCP integration, streaming, and structured outputs.
Tags
Updated: 2026-06-30Capabilities
Typical Inputs
Typical Outputs
What this skill does
- create persistent AI agent
- attach function tools to agent
- attach hosted code interpreter tool
- attach hosted file search tool
- attach hosted web search tool
- integrate MCP server tools
- manage conversation threads
- stream agent responses
- produce structured output responses
- retrieve existing agent by ID
Inputs
- Azure AI project endpoint
- Azure AI model deployment name
- Azure credentials
- Bing connection ID
- MCP server URL
- Function tool definitions
- Pydantic response format model
Outputs
- Agent text responses
- Streaming response chunks
- Structured JSON responses
- Conversation thread IDs
Requirements
- Python runtime environment
- agent-framework or agent-framework-azure-ai package installed
- Azure AI Foundry project provisioned
- Azure credentials (AzureCliCredential or DefaultAzureCredential)
- Environment variables: AZURE_AI_PROJECT_ENDPOINT and AZURE_AI_MODEL_DEPLOYMENT_NAME
- Bing connection configured for web search