发表机构
Intelligent System Department, Zhongxing Telecom Equipment (ZTE); Towngas, China; School of Automation, Northwestern Polytechnical University（中兴通讯智能系统部; 港华燃气; 西北工业大学自动化学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出TopoPlanner，将工具依赖图提升为细胞工作流复形，通过拓扑感知检索与推理为LLM提供规划上下文，在含循环、合并等复杂模式的工作流规划基准上优于现有基线。
AI 中文摘要
面向LLM智能体的任务规划需要满足用户意图和复杂子任务依赖关系的工作流。现有规划器虽能较好地处理顺序或类有向无环图（DAG）结构，但在处理现实工具编排中自然出现的验证-修正循环、汇聚分支合并和可复用中间状态等工作流模式时存在困难。我们提出TopoPlanner，一种拓扑一致的规划框架，它将工具依赖图提升为细胞工作流复形，并将其作为拓扑感知上下文用于LLM工具规划。TopoPlanner通过上同调一致的细胞检索获取与请求相关的闭子复形，在检索到的拓扑上进行多维结构推理，并将得到的细胞表示与规划器LLM对接以生成工具序列。在四个包含拓扑引导的循环、合并及循环-合并工作流的工具规划基准上的实验表明，与基于提示和基于图增强的基线相比，该方法在不同本地LLM骨干网络上均取得了一致的性能提升。
英文摘要
Task planning for LLM agents requires workflows that satisfy both user intent and complex sub-task dependencies. While existing planners work well for sequential or directed acyclic graph (DAG)-like structures, they struggle with workflow patterns such as verification-correction loops, convergent branch merging, and reusable intermediate states that arise naturally in real-world tool orchestration. We present TopoPlanner, a topology-consistent planning framework that lifts tool dependency graphs into cellular workflow complexes and uses them as topologyaware context for LLM tool planning. TopoPlanner retrieves a request-relevant closed subcomplex through cosheaf-consistent cellular retrieval, performs multidimensional structural reasoning over the retrieved topology, and interfaces the resulting cellular representation with the planner LLM for tool-sequence generation. Experiments on four tool-planning benchmarks with topology-guided loop, merge, and loop-merge workflows show consistent improvements over prompt-based and graph-enhanced baselines across different local LLM backbones.
