What Is an Agent Harness?

An agent harness is the software wrapped around a language model that turns it into an agent: the loop that calls the model repeatedly, the tools it can execute, the context management that decides what the model sees, and the guardrails that decide what it’s allowed to do. The model predicts text. The harness is everything that makes those predictions add up to work getting done. Anthropic’s Claude Code documentation puts it in one line: “Claude Code is the harness; Claude is the model inside it.” OpenAI uses the same word for the shared execution core behind every Codex surface (CLI, IDE extension, web), all of them powered by the same Codex harness. When both labs independently settle on a term, it’s worth pinning down what it means. One thing it doesn’t mean: Harness.io, the CI/CD and software delivery company. Same word, coincidental collision, unrelated domain. If you searched “harness devops” and landed here expecting deployment pipelines, that’s them. This post is about the…