Applied AI·Agents
thought, tool call, result, repeat, which is what your loop is already doing whether or not you call it by the paper's name.
ReAct pattern
Draft summary, pending review
The foundational agent loop from a 2022 paper: alternate reasoning ('thought'), action (a tool call) and observation (the result) until done. Modern tool-calling APIs implement its essence natively; the name survives in docs and interviews.