FACET：在终端任务合成中保持源意图与可执行状态

摘要

训练终端智能体需要可扩展的可执行监督，但合成高质量终端任务仍具挑战性。每个任务将指令、初始化环境、参考解和可执行验证器耦合在一起；若这些工件由不一致的假设生成，则所得任务可能不可解或被错误评估。此外，多阶段合成可能丢弃原始来源中蕴含的目标、依赖关系、状态转换和过程约束。我们提出 FACET（可执行任务的细粒度智能体式构建，Fine-grained Agentic Construction of Executable Tasks）框架，同时解决信息保留与跨工件一致性问题。FACET 先将相关智能体技能重建为连贯且信息丰富的场景，再在生成最终任务工件之前实现并修复执行环境。所得容器状态作为指令、解和验证器的共享基础；基于执行的验证与针对性修复可纠正特定工件的失败，而无需不必要地重新生成有效组件。FACET 生成带有密集可执行检查的复杂终端任务，从这些任务中收集的成功轨迹提供了有效且数据高效的监督。对多种规模的模型进行微调持续提升了 Terminal-Bench 2.1 上的性能；对替代生成方案的分析也印证了基于环境的构建对任务有效性和解-验证器对齐的重要性。这些结果确立了源意图保留与共享可执行状态基础作为可扩展终端任务合成的关键原则。

English

Training terminal agents requires scalable executable supervision, yet synthesizing high-quality terminal tasks remains challenging. Each task couples an instruction, an initialized environment, a reference solution, and an executable verifier; if these artifacts are generated from inconsistent assumptions, the resulting task may be unsolvable or incorrectly evaluated. Meanwhile, multi-stage synthesis can discard the goals, dependencies, state transitions, and procedural constraints encoded in the original sources. We present FACET (Fine-grained Agentic Construction of Executable Tasks), a framework that addresses both information preservation and cross-artifact consistency. FACET reconstructs related agent skills into coherent, information-rich scenarios, then realizes and repairs the execution environment before generating the final task artifacts. The resulting container state serves as shared grounding for the instruction, solution, and verifier, while execution-based validation and targeted repair correct artifact-specific failures without unnecessarily regenerating valid components. FACET produces complex terminal tasks with dense executable checks, and successful trajectories collected from these tasks provide effective, data-efficient supervision. Fine-tuning models across multiple scales consistently improves performance on Terminal-Bench 2.1, while analyses of alternative generation schemes support the importance of environment-grounded construction for task validity and solution-verifier alignment. These results establish source-intent preservation and shared executable-state grounding as key principles for scalable terminal-task synthesis.