工具使用LLM代理面临可靠性与成本之间的权衡：将每个决策路由通过LLM可以提高正确性，但会带来高延迟和推理成本，而预编码的工作流图减少了成本，但在意外复合工具故障下变得脆弱。我们提出了Self-Healing Router，一种容错的编排架构，将大多数代理控制流决策视为路由而非推理。该系统结合（i）并行健康监视器，它们为运行时条件（如工具故障和风险信号）分配优先级分数，以及（ii）一个加权工具图，其中Dijkstra算法执行确定性的最短路径路由。当工具在执行中途失败时，其边会被重新加权为无穷大，并重新计算路径——产生自动恢复而无需调用LLM。LLM仅在没有可行路径时才被保留使用，从而实现目标降级或升级。先前基于图的工具使用系统（ControlLLM、ToolNet、NaviAgent）专注于工具选择和规划；我们的贡献是运行时容错，具有确定性的恢复和二进制可观测性——每个故障要么是记录的重新路由，要么是显式的升级，绝不会是沉默的跳过。在19个场景中，覆盖三种图拓扑（线性流水线、依赖DAG、并行扇出），Self-Healing Router在正确性上与ReAct相当，同时将控制平面LLM调用减少93%（9 vs 123汇总），并消除了在复合故障下观察到的静态工作流基线中的沉默故障情况。
英文摘要：
Tool-using LLM agents face a reliability-cost tradeoff: routing every decision through the LLM improves correctness but incurs high latency and inference cost, while pre-coded workflow graphs reduce cost but become brittle under unanticipated compound tool failures. We present Self-Healing Router, a fault-tolerant orchestration architecture that treats most agent control-flow decisions as routing rather than reasoning. The system combines (i) parallel health monitors that assign priority scores to runtime conditions such as tool outages and risk signals, and (ii) a cost-weighted tool graph where Dijkstra's algorithm performs deterministic shortest-path routing. When a tool fails mid-execution, its edges are reweighted to infinity and the path is recomputed -- yielding automatic recovery without invoking the LLM. The LLM is reserved exclusively for cases where no feasible path exists, enabling goal demotion or escalation. Prior graph-based tool-use systems (ControlLLM, ToolNet, NaviAgent) focus on tool selection and planning; our contribution is runtime fault tolerance with deterministic recovery and binary observability -- every failure is either a logged reroute or an explicit escalation, never a silent skip. Across 19 scenarios spanning three graph topologies (linear pipeline, dependency DAG, parallel fan-out), Self-Healing Router matches ReAct's correctness while reducing control-plane LLM calls by 93% (9 vs 123 aggregate) and eliminating the silent-failure cases observed in a well-engineered static workflow baseline under compound failures.
Steering Emotional Dynamics for Art Therapy: Controllable Narrative Script Generation through Hierarchically Guided LLM Agents
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）
Beyond Top-$k$ Skill Retrieval: Diversity-Aware Skill Routing for LLM Agents
Virginia Tech（弗吉尼亚理工大学）；University of Southern California（南加州大学）；Adobe Research（Adobe研究院）；Dolby Labs（杜比实验室）；Texas A&M University（德克萨斯A&M大学）；Arizona State University（亚利桑那州立大学）
ARMOR-MAD: Adaptive Routing for Heterogeneous Multi-Agent Debate in Large Language Model Reasoning
School of Cyber Science and Technology, University of Science and Technology of China（中国科学技术大学网络空间安全学院）；School of Artificial Intelligence, Shenzhen Technology University（深圳技术大学人工智能学院）
MIND: Towards Immersive Psychological Healing with Multi-agent Inner Dialogue
Sichuan University（四川大学）；Shanghai Jiao Tong University（上海交通大学）；Mental Health Center, West China Hospital, Sichuan University（四川大学西昌医院心理健康中心）；WestChina School of Nursing, Sichuan University（四川大学西昌护理学院）；National University of Singapore（新加坡国立大学）
A Threshold-Triggered Deep Q-Network-Based Framework for Self-Healing in Autonomic Software-Defined IIoT-Edge Networks
Energy and Resources Group at the Copernicus Institute of Sustainable Development, Utrecht University（能源与资源组，可持续发展Copernicus研究所，乌特勒支大学）；Operational Technology (OT) System Engineering and Security (Engineering, Procurement, and Construction) at Ørsted Wind Power（运营技术（OT）系统工程与安全（工程、采购与建设）部，Ørsted风能）；System Design Specialists (Engineering, Procurement, and Construction) at Ørsted Wind Power（系统设计专家（工程、采购与建设）部，Ørsted风能）
AgentGate: A Lightweight Structured Routing Engine for the Internet of Agents
School of Artificial Intelligence, University of Science and Technology Beijing（北京科技大学人工智能学院）；China Telecom Research Institute（中国电信研究院）；Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences（中国科学院大学杭州高等研究院）
Modality-Native Routing in Agent-to-Agent Networks: A Multimodal A2A Protocol Extension
AI Architect, Author—Data Engineering for Multimodal AI (O’Reilly)（人工智能架构师，作者—多模态AI的数据工程（O’Reilly））；Stanford School of Engineering (April 2026)（斯坦福大学工程学院（2026年4月））
Efficient and Interpretable Multi-Agent LLM Routing via Ant Colony Optimization
School of Computing, Kyung Hee University（Kyung Hee 大学计算机学院）；School of Information and Software Engineering, University of Electronic Science and Technology of China（中国电子科技大学信息与软件工程学院）；School of Computer Science and Engineering, University of Electronic Science and Technology of China（中国电子科技大学计算机科学与工程学院）；School of Computer Science and Technology, Tongji University（同济大学计算机科学与技术学院）
Beyond Shortest Path: Agentic Vehicular Routing with Semantic Context
Institute of Computing (IC)（计算学院）；Hub de Inteligência Artificial e Arquiteturas Cognitivas (H.IAAC)（人工智能与认知架构中心）；Universidade Estadual de Campinas (UNICAMP)（坎皮纳斯州立大学）
DyTopo: Dynamic Topology Routing for Multi-Agent Reasoning via Semantic Matching
Peking University, Beijing, China（北京大学）；Georgia Institute of Technology, Atlanta, United States（佐治亚理工学院）；Southeast University, Location, Country（东南大学）；Tsinghua University（清华大学）
Orchestrating Intelligence: Confidence-Aware Routing for Efficient Multi-Agent Collaboration across Multi-Scale Models
Research Center for Social Computing and Information Retrieval, Harbin Institute of Technology, China（社会科学与信息检索研究中心，哈尔滨工业大学，中国）；the Institute of Automation of the Chinese Academy of Sciences, China（中国科学院自动化研究所，中国）
