发表机构
Arizona State University; Cisco Research; University of North Carolina at Chapel Hill（亚利桑那州立大学; 思科研究院; 北卡罗来纳大学教堂山分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对多智能体LLM系统持续优化难的问题，提出MASkills框架，通过整合技能条件信用分配等技术优化技能库，在多任务上验证了其有效性。
AI 中文摘要
基于大语言模型（LLM）的多智能体系统在复杂任务上已展现出强大性能，但从交互经验中实现持续优化仍具挑战性。现有自反思方法构建经验记忆，但这些记忆大多难以调用、优化或扩展，而智能体技能提供了更具可操作性的单元：结构化的过程性知识，明确何时行动、如何行动以及使用哪些资源或工具。我们提出MASkills，这是一种通过智能体技能优化多智能体大语言模型系统的持续学习框架。MASkills提出了一种新的智能体优化流程，整合了技能条件信用分配、分层信用聚合和动量平滑优化，使智能体技能库能够通过优化、归纳、整合和剪枝实现演进。在HotpotQA、LoCoMo和GAIA上的实验证明了MASkills在多个智能体任务中的有效性。我们的代码可在该https URL获取。
英文摘要
LLM-based multi-agent systems have shown strong performance on complex tasks, yet continual improvement from interaction experience remains challenging. Existing self-reflection methods build experience memories, but memories are mostly hard to invoke, refine, or scale, while agent skills offer a more actionable unit: structured procedural knowledge that specifies when to act, how to act, and which resources or tools to use. We introduce MASkills, a continual learning framework that optimizes multi-agent LLM systems through agent skills. MASkills presents a new agent-optimization pipeline that integrates skill-conditioned credit assignment, hierarchical credit aggregation, and momentum-smoothed optimization, enabling agent skill libraries to evolve through refinement, induction, consolidation, and pruning. Experiments on HotpotQA, LoCoMo, and GAIA demonstrate the effectiveness of MASkills across multiple agentic tasks. Our code is available at https://github.com/DaRL-GenAI/MASkills
Comments14 pages, 4 figures
Journal refEMNLP 2026 Findings
