Large Language Model (LLM)-based multi-agent systems (MAS) have emerged as a promising paradigm for solving complex tasks. However, existing works often rely on manual designs or "one-size-fits-all" automation, lacking dynamic adaptability after deployment. Inspired by how biological systems adapt, we introduce MASFly, a novel multi-agent framework enabling dynamic adaptation at test time. To adapt system generation, MASFly employs a retrieval-augmented SOP instantiation mechanism that leverages a self-constructed repository of successful collaboration patterns, enabling the LLM to assemble customized MASs for new queries. For adaptive execution, MASFly incorporates an experience-guided supervision mechanism, where a dedicated Watcher agent monitors system behaviors with reference to a personalized experience pool and provides real-time interventions. Extensive experiments demonstrate that MASFly achieves state-of-the-art performance, most notably a 61.7% success rate on the TravelPlanner benchmark, while exhibiting strong task adaptability and robustness.

翻译：基于大型语言模型（LLM）的多智能体系统（MAS）已成为解决复杂任务的一种有前景的范式。然而，现有工作通常依赖于手动设计或“一刀切”的自动化，缺乏部署后的动态适应性。受生物系统自适应方式的启发，我们提出了MASFly，一种新颖的多智能体框架，能够在测试时实现动态自适应。为了自适应地生成系统，MASFly采用了一种检索增强的标准操作程序（SOP）实例化机制，该机制利用一个自构建的成功协作模式知识库，使LLM能够针对新查询组装定制的MAS。为了自适应地执行，MASFly引入了一种经验引导的监督机制，其中一个专门的Watcher智能体参照个性化的经验池监控系统行为，并提供实时干预。大量实验表明，MASFly实现了最先进的性能，尤其是在TravelPlanner基准测试中取得了61.7%的成功率，同时展现出强大的任务适应性和鲁棒性。