The agentic loop is the cycle in which Claude plans, calls a tool, observes the result, and repeats until a stop condition is met.
Each turn the model decides whether to act by calling a tool or to finish; the loop ends when it returns a terminal stop reason or hits a guardrail. Designing reliable loops with clear stop conditions, error handling, and budgets is central to building agents. The Agentic Systems Engineer certification covers this in depth.