# microsoft/agent-framework

A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.

Repository: https://github.com/microsoft/agent-framework
Canonical: https://ross.abutalabs.com/products/agent-framework
Homepage: https://aka.ms/agent-framework
Language: Python
License: MIT
License Family: permissive
Topics: agent-framework, agentic-ai, agents, ai, multi-agent, orchestration, sdk, workflows, dotnet, python
Last push: 2026-08-26T23:53:44+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 35
- inputs: {"age_days": 492, "days_push": 7, "days_rel": 11, "gap_med": 1.0, "n_releases_24m": 111}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 13135, forks 2231 (observed 2026-08-28T04:11:02.261711+00:00)

## 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 in Python, .NET, and Go. It provides graph-based orchestration patterns (sequential, concurrent, handoff, group collaboration) with durability, observability, and provider flexibility across Microsoft Foundry, Azure OpenAI, OpenAI, and GitHub Copilot SDK.

## Use cases
- build production AI agents in Python or .NET
- orchestrate multi-agent workflows with handoff and group chat patterns
- deploy agents with durability, restartability, and human-in-the-loop control
- connect agents to Azure OpenAI, OpenAI, or Microsoft Foundry
- build graph-based agent pipelines beyond a single chat loop
- add observability and governance to agent systems

## When to choose
- you need production-grade agents with orchestration, durability, and observability
- you want multi-language support (Python, .NET, Go) with consistent APIs
- you need provider flexibility across OpenAI, Azure OpenAI, and Microsoft Foundry
- you want graph-based multi-agent patterns like handoff and group collaboration

## When to avoid
- you only need a simple single-prompt or stateless chat loop
- you want a minimal lightweight agent library with no orchestration overhead
- you are locked into a non-Microsoft ecosystem with strict tooling requirements

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, workflow-automation, llm-inference
- domain: large-language-models, artificial-intelligence
- platform: python, dotnet, go, cross-platform
- tags: multi-agent, orchestration, agentic-ai, microsoft, production-grade, human-in-the-loop, observability, ai-agents, automation

## Member repositories
- microsoft/agent-framework (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:02.261711+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:13:30.186107+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/agent-framework (fetched 2026-08-28T04:11:02.261711+00:00, sha 10a4a1dea87e)
  - registry_pypi: https://pypi.org/pypi/agent-framework/json (fetched 2026-08-29T08:08:51.542617+00:00, sha 92e0d5906d93)
- Data as of 2026-08-30T08:39:29.467469+00:00.