How AI Agents Combine Models, Loops, Tools, and Context
Juejin AI presents an AI agent as a model, an execution loop, tools, and context management working together, with the surrounding harness repeatedly invoking the model and returning tool results. It surveys planning patterns, low- and high-code implementation paths, multi-agent orchestration, persistence, and cross-machine collaboration, using WorkBuddy and Claude Code as examples. The article describes Claude Code as using ReAct-style execution with an optional Plan-and-Solve layer, parallel tool calls, and specialized subagents.
The article provides a practical architecture guide for developers building agents, especially for reasoning about where control flow, tool permissions, context, and audit boundaries should live.
Agent 开发认知和架构 & 推理规划和编排
做 Agent,先把一件事说清楚:模型一次只能做一次决策。它能连续查资料、调工具、改方案,是因为外面有一层程序在反复调用它,并把工具结果送回去。这层程序通常叫 Harness。
可以收成一句:
Agent = Model + Loop + Tools + Context Management 模型负责一次决策,Harness 负责让决策持续发生,并守住边界。
下面按这个公式,把落地方式、五种常用的规划模式、大脑系统的几种架构,以及多 Agent 如何协作,串成一条能直接拿去讲的线。例子尽量用同一件事:帮同事安排一次出差,需要查日程、机票、酒店和差旅政策,最后给出一份可执行行程。
1. 最小循环:决策是怎么接起来的
四项各管一块。
#Claude Code#WorkBuddy#DeepSeek V3.2#Dify#Coze#LangChain#LangSmith#A2A