面向长时程终端智能扩展智能体数据

摘要

终端智能体正在超越本地操作，迈向完成真实世界专业任务。短轨迹可以展示命令和工具调用，但无法充分体现智能体如何通过探索积累证据、据此形成判断、根据反馈持续调整策略，并将中间结果转化为最终交付物。高质量长时程轨迹能够在单一问题解决过程中捕捉这些智能体行为模式及其关联，为持续自主工作提供更丰富的经验。然而，环境合成可能提前完成或简化这些模式所依托的关键工作，从而使智能体无需进行其本应执行的探索、实现和验证。为解决这一问题，我们提出 TerminalHorizon，一种面向长时程终端智能体的全自动数据合成引擎。它从真实世界专业工作出发，通过独立执行来识别构建过程中遗漏的关键工作，并利用由此产生的证据指导环境重建。为了将这些模式扩展到不同阶段，它在同一目标下逐步扩展任务，引入以先前结果为基础、并要求进一步探索和改进的新要求。利用该引擎，我们构建了 1.5K 个任务环境，并在所得数据集上对 Qwen3.5-27B 进行监督微调，得到 TerminalHorizon-27B。我们在全部 12 个基准上取得了显著性能提升，这些基准涵盖编码智能体任务、真实生产环境中的智能体任务以及通用推理。多个 RSI 案例研究（每个均设有 12 小时研究预算）进一步表明，该模型能够超越初始解决方案继续实验、根据反馈改进方法，并在最终交付物中保留有效的改进。

English

Terminal agents are moving beyond local operations toward completing real-world professional tasks. Short trajectories can demonstrate commands and tool calls, but they do not fully show how agents accumulate evidence through exploration, use it to inform judgments, continually adjust strategies based on feedback, and turn intermediate results into final deliverables. High-quality long-horizon trajectories capture these agentic behavioral patterns and their connections within a single problem-solving process, providing richer experience for sustained autonomous work. However, environment synthesis may complete or simplify the key work underlying these patterns in advance and thus remove the need for exploration, implementation, and verification that the agent was meant to perform. To address this, we introduce TerminalHorizon, a fully automated data synthesis engine for long-horizon terminal agents. Starting from real-world professional work, it uses independent execution to identify key work omitted during construction and uses the resulting evidence to guide environment reconstruction. To extend these patterns across stages, it progressively expands tasks under the same goal, introducing new requirements that build on prior results and call for further exploration and improvement. Using this engine, we construct 1.5K task environments and perform supervised fine-tuning of Qwen3.5-27B on the resulting dataset to obtain TerminalHorizon-27B. We achieve substantial performance gains on all 12 benchmarks spanning coding-agent tasks, agentic tasks in real-world production settings, and general reasoning. Multiple RSI case studies, each with a 12-hour research budget, further demonstrate the model's ability to continue experimenting beyond an initial solution, improve its methods based on feedback, and retain effective improvements in its final deliverables.

面向长时程终端智能扩展智能体数据

Scaling Agentic Data for Long-Horizon Terminal Intelligence

摘要

Support