用自然语言描述智能体执行流程，让任务调度更透明可比。
Natural-Language Agent Harnesses
- 用可编辑的自然语言文档定义任务执行策略
- 在多个任务上表现接近代码实现，但策略更简短
- 适合研究者分析和对比不同任务调度逻辑
智能体性能受其外部执行系统（即任务运行的调度框架）强烈影响，但传统做法将此逻辑嵌入紧密耦合的控制器代码中，导致难以观察、比较、迁移或消融。本文提出将智能体调度模式抽象为可执行的自然语言对象——自然语言智能体调度器（NLAH），并设计智能调度运行时（IHR），能解析这些文档，生成代理调用、状态更新、验证节点与成果契约。在编程、终端使用和计算机操作等基准测试中，IHR执行的NLAH实现与代码和提示式实现相当的任务效果，同时具备更短的静态调度策略。模块消融实验进一步表明，显式调度模块具有可分析性。结果表明，智能体调度可从模型周围的附加工具转变为可科学研究的表达对象。
原文摘要 · Abstract (English)
Agent performance is strongly shaped by the surrounding harness: the external execution system around a model that organizes a task run. Yet this logic is usually buried in tightly coupled controller code, which makes harnesses hard to inspect, compare, transfer, and ablate. This paper asks whether the reusable design pattern of an agent harness can be represented as an executable natural-language object. We introduce Natural-Language Agent Harnesses (NLAHs), editable documents that describe run-level harness policy, and Intelligent Harness Runtime (IHR), a shared runtime that interprets these documents into agent calls, handoffs, state updates, validation gates, and artifact contracts. Across coding, terminal-use, and computer-use benchmarks, IHR-executed NLAHs achieve comparable task outcomes to code and prompted realizations, while exposing much shorter static harness policies. Module ablations further show that explicit harness modules are analyzable. These results suggest that agent harnesses can be turned from incidental glue around models into scientific representation objects.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。