GoKawiilAnthropic defines 'agent,' 'harness,' and 'model' as distinct layers of AI systems
Anthropic published a terminology guide clarifying that an 'agent system' is not the same as a 'model.' It breaks the stack into three parts: the model (like Sonnet or Opus), the inference service that runs it (like AWS Bedrock or Anthropic's API), and the harness, the interface layer such as Claude Desktop or Claude CLI that manages tools and context.
GoKawiil's interpretation of the reporting above, not reported fact.
Features like MCP and Skills live in the harness layer, not the model itself, meaning the underlying model has no inherent awareness of these tools unless the harness exposes them. This distinction matters for developers building on Claude, since confusing these layers can lead to misunderstandings about what capabilities are model-driven versus interface-driven.
- Anthropic separates AI systems into three layers: model, inference service, and harness.
- MCP and Skills operate at the harness layer, not within the model itself.
- Clear terminology helps developers reason precisely about how agent systems are built and behave.
O'Reilly "AI Engineering" by Chip Huyen — If this breakdown of models, inference services, and agent harnesses clicked for you, Chip Huyen's AI Engineering goes deeper on exactly those layers, from serving foundation models to building applications on top of them. It's a solid desk reference for getting your terminology and architecture right when talking about agent systems.
See O'Reilly "AI Engineering" by Chip Huyen on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: code.joejag.com, 2026-08-24
Published there as: “Agent Is Not the Model”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.