基於技能圖譜的可擴展終端任務生成研究

摘要

终端智能体已展现出强大的自主命令行执行潜力，但其训练仍受限于高质量多样化执行轨迹的稀缺性。现有方法通过合成大规模终端任务实例进行轨迹采样以缓解这一瓶颈，但主要侧重于任务数量的扩展，对智能体实际训练过程中执行轨迹多样性的控制能力有限。本文提出SkillSynth——一种基于场景中介技能图的终端任务自动合成框架。该方法首先构建大规模技能图，以场景作为中间过渡节点连接多样化命令行技能；随后从图中采样路径作为现实工作流的抽象表示，并通过多智能体系统将其实例化为可执行任务。通过基于图采样工作流路径的任务合成机制，SkillSynth能显式控制解决合成任务所需最小执行轨迹的多样性。在Terminal-Bench上的实验验证了该框架的有效性。此外，Hy3 Preview已采用SkillSynth合成的任务实例进行训练，显著提升了其在终端环境中的智能体能力。

English

Terminal agents have demonstrated strong potential for autonomous command-line execution, yet their training remains constrained by the scarcity of high-quality and diverse execution trajectories. Existing approaches mitigate this bottleneck by synthesizing large-scale terminal task instances for trajectory sampling. However, they primarily focus on scaling the number of tasks while providing limited control over the diversity of execution trajectories that agents actually experience during training. In this paper, we present SkillSynth, an automated framework for terminal task synthesis built on a scenario-mediated skill graph. SkillSynth first constructs a large-scale skill graph, where scenarios serve as intermediate transition nodes that connect diverse command-line skills. It then samples paths from this graph as abstractions of real-world workflows, and uses a multi-agent harness to instantiate them into executable task instances. By grounding task synthesis in graph-sampled workflow paths, SkillSynth explicitly controls the diversity of minimal execution trajectories required to solve the synthesized tasks. Experiments on Terminal-Bench demonstrate the effectiveness of SkillSynth. Moreover, task instances synthesized by SkillSynth have been adopted to train Hy3 Preview, contributing to its enhanced agentic capabilities in terminal-based settings.

基於技能圖譜的可擴展終端任務生成研究

Toward Scalable Terminal Task Synthesis via Skill Graphs

摘要

Support