Independent developer guide · August 3, 2026

Choose the smallest Agent Framework runtime that fits the task

Microsoft Agent Framework is an open-source SDK for agents and multi-agent workflows. Its agents, Harness, and workflows share foundations, but they solve different control problems—and none of them automatically grants safe data or tool authority.

Architecture decision

Agents, Harness, workflows, and functions are different operating shapes

Agents

Open-ended decisions, conversation, and autonomous tool use

Choose the model, session, context providers, tools, middleware, MCP clients, and approval policy.

Harness

One agent carrying a long task across many service calls

Stable Python and .NET core; verify the separate lifecycle and permissions of optional file, shell, loop, and background features.

Workflows

Known graphs with routing, checkpoints, functions, agents, and human review

Use programmatic graphs for custom logic or declarative YAML for standard, reviewable sequences.

Function

A deterministic operation with known inputs and outputs

Prefer ordinary code when it is sufficient; an agent adds cost, uncertainty, and a new authority surface.

SDK support

Language support is not feature parity

Python

Primary SDK

Install from the reviewed Agent Framework packages and pin the exact set used by agents, Harness, workflows, and providers.

.NET

Primary SDK

Current Learn examples can use prerelease package flags even where a component has a stable announcement; verify each artifact.

Go

Public preview

The overview currently excludes declarative agents, RAG, CodeAct, and functional workflows from Go support.

Package feeds, provider integrations, adapters, and optional capabilities can have narrower lifecycle labels than the core SDK. Lock and test the exact dependency graph.

Production path

Prove task fit, authority, and recovery before adding autonomy

01

Choose a function, workflow, agent, Harness, or Squad from the task shape—not from feature breadth.

02

Pin SDK, provider, adapter, model, and schema versions; record preview flags and runtime warnings.

03

Separate data access from action authority, and scope file, shell, URL, MCP, and state-changing tools independently.

04

Set hard budgets for calls, tokens, elapsed time, loops, retries, and external spend outside model instructions.

05

Evaluate fixed cases, adversarial inputs, denied tools, retries, cancellation, resume, and target-system receipts.

Microsoft Agent Framework questions, answered

Microsoft Agent Framework is an open-source SDK for building AI agents and multi-agent workflows in Python, .NET, and preview Go. It combines agent abstractions, model clients, sessions, context providers, middleware, MCP clients, an opinionated Harness, and graph-based workflows.
Microsoft describes Agent Framework as the successor that combines concepts from AutoGen and Semantic Kernel. Migration still requires an API, behavior, provider, storage, tool, and evaluation review rather than a package-name replacement.
Use an agent for open-ended decisions, conversation, and dynamic tool use. Use a workflow for known steps, typed routing, checkpoints, multiple agents or functions, and human review. If a normal function is sufficient, prefer the function.
No. Application builders remain responsible for data boundaries, credentials, tool schemas, approvals, model and provider behavior, third-party terms, logging, evaluation, and the effect of external actions.
No. This is an independent, source-checked developer guide. It does not install the SDK, grant Microsoft or GitHub access, execute tools, or establish a Flowith integration.

This page is an independent information guide, not an embedded Microsoft runtime or a Flowith integration. Source check: August 3, 2026. Verify current SDKs, packages, providers, lifecycle labels, model support, storage, tools, approvals, telemetry, third-party terms, and deployment controls.