The Agent Loop — what's actually running when an agent "works for hours"featured
An AI agent is a loop, not a magic box. This course teaches you to draw that loop from memory, name its five components, and pattern-match the failure modes that show up between hour one and hour six — so you can spec an autonomous system without hand-waving.
Lessons
Foundations
01What an agent actually is — the loop, drawn7 min02The context window as working memory8 min03The tool-call surface — the verbs the agent can speak7 min04Stopping conditions and budget guards8 minFailure modes
05Failure modes between hour one and hour six8 min