https://piotrgoral.github.io/tags/agents/Recent content in Agents on AI from first principlesHugoen-usSun, 13 Sep 2026 00:00:00 +0000https://piotrgoral.github.io/posts/what-is-a-harness/Sun, 13 Sep 2026 00:00:00 +0000https://piotrgoral.github.io/posts/what-is-a-harness/<h2 id="in-short">In short</h2> <p>This article builds a framework that defines what a Harness is and where its boundaries lie, based on existing definitions. The framework separates the layers of an agent system: prompt, context, LLM agent, and Harness. Within the Harness, it defines a core of three parts: environment, turn lifecycle, and verification. Beyond that core, the framework can be extended with cross-session memory, interfaces, and observability. Additionally, a self-improving harness and the loop can operate on the framework. The framework is then validated against existing harness implementations (coding agents, harness SDKs, and personal assistants) by analyzing their code.</p>