[{"data":1,"prerenderedAt":74},["ShallowReactive",2],{"term-r\u002Freact-pattern":3,"related-r\u002Freact-pattern":59},{"id":4,"title":5,"acronym":6,"body":7,"category":40,"description":41,"difficulty":42,"extension":43,"letter":44,"meta":45,"navigation":46,"path":47,"related":48,"seo":53,"sitemap":54,"stem":57,"subcategory":6,"__hash__":58},"terms\u002Fterms\u002Fr\u002Freact-pattern.md","ReAct Pattern",null,{"type":8,"value":9,"toc":33},"minimark",[10,15,19,23,26,30],[11,12,14],"h2",{"id":13},"eli5-the-vibe-check","ELI5 - The Vibe Check",[16,17,18],"p",{},"ReAct is thinking out loud while doing chores. 'I think the milk is in the fridge, let me check, yep it's here, now I'll grab it.' Instead of silently deciding what to do or blindly acting without a plan, the model narrates a thought, takes an action, reads what happened, and narrates the next thought. That narration is what makes its behavior debuggable instead of a black box.",[11,20,22],{"id":21},"real-talk","Real Talk",[16,24,25],{},"ReAct (Reasoning and Acting) is a prompting pattern where the model alternates between generating a short reasoning trace and taking an action, typically a tool call, then observing the result before producing the next reasoning step. Introduced in a 2022 Google Research paper, it's the template most modern tool-using agent loops are built on, even when the reasoning trace itself isn't shown to the user.",[11,27,29],{"id":28},"when-youll-hear-this","When You'll Hear This",[16,31,32],{},"\"The agent's ReAct trace shows it second-guessed itself twice before calling the right tool.\" \u002F \"Most tool-calling agents are just ReAct with better prompts.\"",{"title":34,"searchDepth":35,"depth":35,"links":36},"",2,[37,38,39],{"id":13,"depth":35,"text":14},{"id":21,"depth":35,"text":22},{"id":28,"depth":35,"text":29},"ai","ReAct is thinking out loud while doing chores. 'I think the milk is in the fridge, let me check, yep it's here, now I'll grab it.","advanced","md","r",{},true,"\u002Fterms\u002Fr\u002Freact-pattern",[49,50,51,52],"Agent Loop","Chain of Thought","Tool Use","Function Calling",{"title":5,"description":41},{"changefreq":55,"priority":56},"weekly",0.7,"terms\u002Fr\u002Freact-pattern","UPnfQshYOEYlQnBdG3dvI_mj8bWvJklF_dLozDloOC8",[60,64,68,71],{"title":49,"path":61,"acronym":6,"category":40,"difficulty":62,"description":63},"\u002Fterms\u002Fa\u002Fagent-loop","intermediate","The agent loop is the heartbeat of any AI coding agent: think, act, look, repeat.",{"title":50,"path":65,"acronym":66,"category":40,"difficulty":62,"description":67},"\u002Fterms\u002Fc\u002Fchain-of-thought","CoT","Chain of Thought is when you tell the AI 'show your work' like a math teacher.",{"title":52,"path":69,"acronym":6,"category":40,"difficulty":62,"description":70},"\u002Fterms\u002Ff\u002Ffunction-calling","Function Calling is the OpenAI term for what Anthropic calls Tool Use — teaching the AI to call your code functions.",{"title":51,"path":72,"acronym":6,"category":40,"difficulty":62,"description":73},"\u002Fterms\u002Ft\u002Ftool-use","Tool use is when an AI can call external functions, APIs, or programs to do things it can't do alone.",1792528114810]