Microsoft Agent Framework

Microsoft Agent Framework is the open‑source framework for the developers and AI engineers building production agentic applications — and the direct successor to both Semantic Kernel and AutoGen, created by the same teams. Two primary capabilities anchor the framework: Agents, which use LLMs to process inputs, call tools and MCP servers, and generate responses across providers including Microsoft Foundry, Anthropic, Azure OpenAI, OpenAI, and Ollama; and Workflows, graph‑based pipelines that connect agents and functions for multi‑step tasks with type‑safe routing, checkpointing, and human‑in‑the‑loop support. Put simply, Agent Framework merges AutoGen’s simple agent abstractions with Semantic Kernel’s enterprise‑grade state management, type safety, middleware, and telemetry — and adds explicit multi‑agent orchestration on top.

This is the framework you reach for when a single agent is no longer enough. Builders shipping conversational copilots and tool‑using assistants get an agent abstraction that’s simple to start but doesn’t fall apart as scope grows. Teams composing multiple agents — researcher plus writer plus reviewer, planner plus executors, agentic data pipelines — get explicit graph‑based workflows with checkpointing and human‑in‑the‑loop control instead of brittle ad‑hoc prompt chains. And because Agent Framework is open source on GitHub, the same code runs from prototype to production, across providers and across deployment targets.

Start building agents today — try Microsoft Agent Framework on GitHub.