// ⌘ Agents & Tool Use

ReAct (Reason + Act)

Interleaves reasoning, actions and observations.

ReAct 推理-行动循环 2022 Foundational core

IN PLAIN WORDS

The model writes a thought, then an action calling a tool, reads the observation and continues until done. This loop underlies nearly every agent framework.

Used in: LangChain/LangGraph, coding and browser agents

TIMELINE · prerequisites → this → what builds on it

2022 Chain-of-Thou…2022 ReAct (Reason + A…2023 Function Call…2023 Reflexion2023 Generative Ag…2023 Multi-Agent C…2024 Agent–Compute…

01Original paper

ReAct: Synergizing Reasoning and Acting in Language Models

Shunyu Yao, Jeffrey Zhao, Dian Yu, et al. · 2022

arXiv ↗ ✓ Verified · arXiv API

02Where it sits

Learn the left column first; the right column builds on this one.

Learn first

Chain-of-Thought Prompting

ReAct (Reason + Act)

Then you can learn

Related:Model Context Protocol (MCP)

Left: prerequisites · right: builds on this · dashed: related. Full map →

03Learning resources

Official docs, explainers, courses and reference code. All links checked.

Official docs

react-lm.github.io ↗EN

Explainers

04Where the course teaches it

U82 Agents: Tool Use, MCP, Planning & Memory ReAct, function calling, MCP, reflection, memory, multi-agent systems, computer use and sandboxes. 12 h10 algorithms4 paperslesson ready

On learning paths:AI Application Engineer

For AI / data

Verification: verified, checked 2026-10-21.
JSON: /api/learn/algorithms.json (id: react)