What an agent actually is

A first-principles definition of an AI agent: control loops, tools, state, and decisions, and what agent frameworks are really doing underneath their abstractions.

The series replaces the “Agent” definition question with an autonomy spectrum, then opens up what frameworks actually do underneath: the model emits, the API parses, your code dispatches and loops. The point is to know what kind of agent you have, and how it actually works.

Read in order: