Cognition 模型 / Devin 博客·精选AI 评分68
Cognition 发文提出上下文工程原则，称多智能体架构在 2025 年仍易碎
Don’t Build Multi-Agents2025.06.12Frameworks for LLM Agents have been surprisingly disappointing. I want to offer some principles for building agents based on our own trial & error, and explain why some tempting ideas are actually quite bad in practice.
AI 导读
Cognition（Devin 团队）发文 Don't Build Multi-Agents，基于实战提出两条构建 Agent 的原则：共享完整 Agent 轨迹的上下文，以及行动隐含决策、冲突决策带来坏结果。
推荐理由
作者基于 Devin 的实战经验提出上下文工程两条原则，并解释为何多智能体架构在实际中易碎。
来源：Cognition 模型 / Devin 博客 · cognition.com