AI pills Agent Architecture Deep-Dive: Reflexion Two weeks ago I covered ReAct, the foundational reason → act → observe loop that most agents are built on. ReAct reasons well within a single pass, but it has no mechanism for actually getting better when that pass fails. Reflexion fixes that gap. How it works Reflexion wraps a task attempt
ReAct ReAct: The Loop Behind Most AI Agents When people say they've "built an AI agent," they usually mean the model can call tools and decide when to call them. Most of those systems are quietly running on a pattern called ReAct, short for Reason + Act. It's worth knowing on purpose, because
