Agentic Loop
The cycle an AI agent runs: think, pick a tool, observe the result, decide whether to continue.
- Alternates reasoning with action; variants include ReAct, Plan-and-Execute.
- Ends on declared done or step-limit.
- Stuck loop burns tokens; wrong early step compounds.
When to use: Whenever multi-step tool use is needed; with budgets and safeguards.
Example: "Fix failing test": read test, read source, edit source, run test, see error, edit again, run again, done.