Agent decision-making where the environment state evolves during reasoning, requiring fast responses.