agent-framework
A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.
What is agent-framework?
A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.
Overview of agent-framework
What it is
Microsoft Agent Framework (MAF) is an open-source, multi-language framework for building, orchestrating, and deploying production-grade AI agents and multi-agent workflows. It supports Python and .NET (C#) with consistent APIs, and also provides a separate Go SDK. The framework is designed for teams that need to move agents from prototype to production, offering a consistent foundation across languages while keeping architecture choices flexible.
Key features
- Python and C#/.NET Support: Full framework support with consistent APIs for both Python and .NET implementations.
- Go Support: A separate Go SDK is available at microsoft/agent-framework-go.
- Multiple Agent Provider Support: Support for various LLM providers, with more being added continuously.
- Middleware: Flexible middleware system for request/response processing, exception handling, and custom pipelines.
- Orchestration Patterns & Workflows: Graph-based workflows supporting sequential, concurrent, handoff, and group collaboration patterns; includes checkpointing, streaming, human-in-the-loop, and time-travel.
- Foundry Hosted Agents: Deploy and host agents to Foundry-hosted infrastructure with just 2 additional lines of code.
- Observability: Built-in OpenTelemetry integration for distributed tracing, monitoring, and debugging.
- Declarative Agents: Define agents using YAML for faster setup and versioning.
- Agent Skills: Build domain-specific knowledge bases from multiple sources for agents to discover and use.
- AF Labs: Experimental packages for benchmarking, reinforcement learning, and research initiatives.
- DevUI: Interactive developer UI for agent development, testing, and debugging workflows.
Use cases
- Building agents and workflows expected to run in production.
- Scenarios requiring orchestration beyond a single prompt or stateless chat loop.
- Applications needing graph-based patterns such as sequential, concurrent, handoff, and group collaboration.
- Systems that require durability, restartability, observability, governance, or human-in-the-loop control.
- Projects needing provider flexibility so the architecture can evolve without major rewrites.
Getting started
Installation commands:
### Python
pip install agent-framework
### .NET
dotnet add package Microsoft.Agents.AI
### For Foundry integration:
dotnet add package Microsoft.Agents.AI.Foundry
dotnet add package Azure.AI.Projects
dotnet add package Azure.Identity
For the Go SDK, see microsoft/agent-framework-go. Quickstart code examples for Python and .NET are available in the repository README.
License and pricing
License: MIT. The project is open source. No pricing information is stated beyond being open source.
Based on the project's public README and repository metadata.