See it, don't just read it
Every core idea comes with a demo you step through in the browser — no setup, no API keys, no code to run. Press a button, watch the loop turn.
The Harness
Everyone talks about AI agents. Almost nobody explains the machinery that makes one survive contact with the real world. This does — from the ground up, with demos you can poke.
What this is
Section titled “What this is”An AI agent is not magic and it is not a chatbot. It is a plain loop around a language model, wrapped in a layer of infrastructure that most tutorials skip entirely. That layer is the harness — and it is the difference between a demo that works on stage and a system that works on a Tuesday afternoon when the process crashes mid-task.
This guide builds that understanding from zero. No prior agent experience assumed.
Built for mixed rooms
Whether you write code or lead the people who do, you can follow the whole story. Code is shown to make ideas concrete, never as a prerequisite.
Concepts before jargon
We earn every term. “Durable execution,” “context hydration,” “supervision” — each one arrives only after you’ve felt the problem it solves.
Ends at the real world
By the end you can read the docs for LangGraph, Temporal, or any agent framework and recognize exactly what each piece is doing, and why.
The one idea to hold onto
Section titled “The one idea to hold onto”Agent systems are workflow systems. The model decides what to do next. The harness decides how it actually gets done — safely, durably, and without losing its mind halfway through.
Everything in this guide is an elaboration of that single sentence.