agent framework
In plain wordsA software library that gives developers ready-made parts for building AI agents — model connections, tool handling, memory, multi-agent coordination and logging — so they need not write everything from scratch.
A software library or SDK providing abstractions for building agentsAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done. and agentic workflowsagentic workflow, n.: A multi-step process in which AI models draft, check, use tools and revise their work over several rounds, instead of answering in one shot; often the steps are laid out by developers in advance.: adapters for model providers, tool definition and execution, the agent loopagent loop, n.: The repeating cycle at the heart of an AI agent: the model decides what to do next, software carries it out, the result is shown to the model, and the cycle repeats until the task is finished., memory and state persistence, multi-agent orchestrationorchestration, n.: Coordinating several AI calls, tools or agents so each does its part in the right order and their results are combined into one finished piece of work, as a conductor directs an orchestra. and handoffshandoff, n.: The moment one AI agent passes a conversation or task to another, better-suited agent — like a call-center operator transferring you to the right department., guardrails, human-approval steps, tracing and evaluation. Frameworks differ in their central abstraction — chains and graphs (LangChain, LangGraph), conversing agents (Microsoft's AutoGen and its successor, Microsoft Agent Framework), role-based crews (CrewAI), handoffs and guardrails (OpenAI Agents SDK), code-writing agents (Hugging Face's smolagents), or a production harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car. exposed as a library (Anthropic's Claude Agent SDK) — and in how much control flow they hide. A recurring criticism is that heavy abstraction obscures the prompts and context actually sent to the model, so many practitioners prefer thin libraries or direct API calls; Anthropic's 2024 guidance advises starting without a framework.
We dropped the agent framework and wrote the loop ourselves in eighty lines; debugging got much easier.
Origin
LangChainLangChain, prop. n.: A popular free software toolkit, and the company behind it, for building applications that connect AI language models to documents, databases, tools and multi-step workflows. (Harrison Chase, October 2022) and LlamaIndexLlamaIndex, prop. n.: A free software toolkit, and the company behind it, for connecting AI language models to an organization's own documents and data so the models can find and use that information. (2022) were the first widely used frameworks; AutoGPTAutoGPT, prop. n.: A 2023 open-source program that gave GPT-4 a goal and let it run on its own — planning, searching the web and writing files in a loop — becoming the first viral 'autonomous AI agent'., Microsoft's AutoGen (2023) and CrewAI (2023) followed. Model developers entered with OpenAI's Agents SDK (March 2025), Google's Agent Development Kit (April 2025) and Anthropic's Claude Agent SDK (first released as the Claude Code SDK and renamed in September 2025).
Usage note
Not the same as a harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car.: a framework is a toolkit for building agents, while a harness is the runtime of a particular agent. The line blurs when a harness is published as an SDK.
Related
- harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car.
- LangChainLangChain, prop. n.: A popular free software toolkit, and the company behind it, for building applications that connect AI language models to documents, databases, tools and multi-step workflows.
- orchestrationorchestration, n.: Coordinating several AI calls, tools or agents so each does its part in the right order and their results are combined into one finished piece of work, as a conductor directs an orchestra.
- multi-agent systemmulti-agent system, n.: A setup in which several AI agents, each with its own role or point of view, work together — or compete — on a problem, passing messages and results among themselves.
- handoffhandoff, n.: The moment one AI agent passes a conversation or task to another, better-suited agent — like a call-center operator transferring you to the right department.
- DSPyDSPy, prop. n.: A free software framework in which programmers describe what each step of an AI pipeline should do, and the framework automatically works out good prompts and examples instead of hand-writing them.
- LlamaIndexLlamaIndex, prop. n.: A free software toolkit, and the company behind it, for connecting AI language models to an organization's own documents and data so the models can find and use that information.
- harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car.
Sources
- OpenAI (2025). New tools for building agents. 11 March 2025. openai.com/index/new-tools-for-building-agents
- Schluntz, E. & Zhang, B. (2024). Building effective agents. Anthropic. anthropic.com/research/building-effective-agents
Mentioned in
AutoGPTAutoGPT, prop. n.: A 2023 open-source program that gave GPT-4 a goal and let it run on its own — planning, searching the web and writing files in a loop — becoming the first viral 'autonomous AI agent'., harnessharness, n.: The software around an AI model that gives it instructions and tools, carries out its actions and decides when it stops. The model is the engine; the harness is the rest of the car., LangChainLangChain, prop. n.: A popular free software toolkit, and the company behind it, for building applications that connect AI language models to documents, databases, tools and multi-step workflows., orchestrationorchestration, n.: Coordinating several AI calls, tools or agents so each does its part in the right order and their results are combined into one finished piece of work, as a conductor directs an orchestra., scaffoldingscaffolding, n.: The code, prompts and tools built around an AI model so it can carry out tasks it could not manage from a bare prompt — like the scaffolding that lets builders reach every part of a building., tool calltool call, n.: A single request from an AI model to use one of its tools — naming the tool and filling in the details — which the surrounding software then carries out and reports back on., AutoGenAutoGen, prop. n.: A free Microsoft software framework for building teams of AI agents that solve problems by messaging one another, writing and running code, and asking people for help when needed., CrewAICrewAI, prop. n.: A free software framework for setting up teams of AI agents, each given a job title, goal and backstory, that divide up tasks and work through them together like a small staff., handoffhandoff, n.: The moment one AI agent passes a conversation or task to another, better-suited agent — like a call-center operator transferring you to the right department., LangGraphLangGraph, prop. n.: A free software library for building AI agents as flowcharts of steps that share a memory, so developers control exactly how an agent moves between steps, pauses and resumes., LlamaIndexLlamaIndex, prop. n.: A free software toolkit, and the company behind it, for connecting AI language models to an organization's own documents and data so the models can find and use that information., MRKL systemMRKL system, n.: An early design, pronounced 'miracle', in which a language model acts as a switchboard, sending each part of a question to the right specialist — a calculator, a database, a search tool — and combining the answers.