An agent is a model calling tools in a loop until it reaches a goal: it reads the situation, picks an action, observes the result, and repeats. Understanding the loop shows where controls belong, such as before a tool runs, after a result comes back, and when the loop should stop and ask a human.
Read in this order
- ArticleBuilding effective AI agents (Anthropic)Start here: the core agent patterns, and when a simple workflow is enough.
- ArticleHow to Build an Agent (Thorsten Ball, Amp)Build a working agent in a few lines of code and see the loop run.
- VideoWhat Are AI Agents? Agentic Systems Explained | Stanford CS329Z Lecture 1 with Diyi YangDistinguishes an agent from a workflow before teaching controls. 70 min.