MAS-on-the-Fly: In-Context Structural Adaptation of LLM-Based Multi-Agent Systems

Large Language Model (LLM)-based multiagent systems (MAS) have emerged as a promising paradigm for complex tasks.However, existing works often rely on manual designs or "one-size-fits-all" automation and lack adaptability after deployment.We study incontext structural adaptation, where structured experience conditions both query-dependent system generation and execution-time reconfiguration without updating LLM parameters.We introduce MASFly, which realizes this adaptation through two complementary mechanisms.First, a retrieval-augmented SOP instantiation mechanism retrieves and adapts successful collaboration patterns to construct a query-specific MAS.Second, an experience-enhanced process supervision mechanism uses a dedicated Watcher agent to monitor execution against prior failure experience and reconfigure the system upon abnormal behavior.Experiments demonstrate that MASFly achieves state-ofthe-art performance, including a 61.7% success rate on TravelPlanner, with strong task adaptability and robustness.Our code and data are available at https://github.com/ LARS-research/MASFly.