Agent harness

The execution system that wraps a model and turns it into a working agent - owning the agent loop, conversation state, the permission policy, the sandbox and recovery across turns.

A harness is distinct from a framework, which supplies primitives you compose, and from a protocol, which is a wire contract that runs nothing. Claude Agent SDK, the OpenAI Agents API and Claude Code are harnesses; LangGraph and the OpenAI Agents SDK are frameworks. Choosing between them is choosing whether to configure an opinionated loop or compose your own.

Explained in depth

Agent harnesses - the layer that owns the loop

The difference between an agent harness, an agent framework, and a protocol - what a harness actually provides (loop, permissions, sandbox, context, recovery) and why the layer names get confused.

Agent sandboxes - where unvetted code runs

The infrastructure layer that runs code an agent wrote without a human reviewing it - the isolation primitives (container, gVisor, microVM), the lifecycle a sandbox exposes, and how E2B, Modal and Daytona differ on the dimensions that actually break a deployment.

Agent Skills - packaged instructions with progressive disclosure

What an Agent Skill actually is - a folder with a SKILL.md file, a three-level progressive-disclosure model that controls what enters context and when, and why the description field is the whole trigger mechanism.

Learn it in context