#3737th of 48 in Agent frameworks and SDKs
agent-framework
Microsoft's Python and .NET framework for agents and multi-agent workflows
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 14k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
Microsoft Agent Framework is a code-first library for building agents and graph-based multi-agent workflows in Python and .NET. Workflows support sequential, concurrent, handoff and group patterns, with checkpointing, streaming and human-in-the-loop. It includes middleware, OpenTelemetry tracing, YAML declarative agents, and a DevUI for testing. It is positioned as the successor path for Semantic Kernel and AutoGen users.
Who it is for: .NET and Python teams taking multi-agent systems to production
Strengths
- Same agent and workflow APIs in Python and .NET, with a separate Go SDK
- Graph workflows with checkpointing, streaming, human-in-the-loop and time-travel
- Built-in OpenTelemetry tracing and a middleware pipeline
- Migration guides from Semantic Kernel and AutoGen
Weaknesses
- Quickstarts target Microsoft Foundry and Azure credentials; other providers need extra setup
- Durable Task and Azure Functions hosting live in a separate extension repo
- Experimental features sit in a separate lab package
- Go SDK is maintained in a different repository
What it needs
- no GPU
- Needs Microsoft Foundry or Azure OpenAI or OpenAI (LLM provider), Azure CLI (for az login in quickstart)
- Models: Microsoft Foundry, Azure OpenAI, OpenAI, GitHub Copilot SDK
Also in Agent frameworks and SDKs
See all 48| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 93 out of 100 |
| 2 | browser-usePython library that lets an LLM agent control a web browser | 89 out of 100 |
| 3 | langchainPython framework for building LLM agents and apps from interoperable components | 88 out of 100 |
| #4 | tradingagentsLangGraph multi-agent framework simulating a trading firm to analyze tickers | 88 out of 100 |
| #5 | copilotkitSDK for adding agent chat, generative UI and shared state to apps | 87 out of 100 |