ReAct
Reasoning and Acting is a prompt engineering paradigm that forces an LLM to generate explicit internal thoughts before taking an action with an external tool.
Think of It Like This
Like a detective thinking out loud, 'The safe is locked, so I need a key,' before actually reaching out to grab the key from the desk.
In a ReAct loop, the model outputs a 'Thought' analyzing its current state, an 'Action' requesting a tool use (like a web search), and receives an 'Observation'. This structured cycle significantly improves an agent's ability to solve complex, multi-step problems. It prevents the model from hallucinating immediate answers without verifying facts first.