将智能体成功轨迹提炼为可复用的程序化技能，提升任务执行效率。
SKILL-DISCO: Distilling and Compiling Agent Traces into Reusable Procedural Skills
- 从成功轨迹中提取可复用的参数化控制流子图作为技能
- 在ALFWorld和WebArena上提升成功率并减少执行轮次
- 适合需要长期记忆与重复执行的任务场景
智能体在解决相似任务时往往从头开始推理，造成不必要的计算开销和长执行轨迹。已有工作探索过工作流复用和可执行技能生成，但尚未明确哪些任务场景适合生成程序化技能，以及如何统一表示跨成功轨迹的共享结构。本文研究有限状态机（FSM）定义的场景，将成功轨迹视为未知转移图中的路径，将程序化技能形式化为可复用的参数化控制流子图。基于此，提出SkillDisCo框架，通过蒸馏和编译，从成功轨迹中提取并生成可调用、可执行、可验证的程序化技能。在ALFWorld和WebArena上的实验表明，SkillDisCo在多个基准和模型规模下均提升了成功率并减少了智能体回合数，验证了以可复用执行结构表示共享经验的有效性。
原文摘要 · Abstract (English)
Agents often repeatedly solve similar task instances from scratch, leading to unnecessary reasoning cost and long execution traces. Prior work has explored workflow reuse and executable skill induction, but it remains unclear which task scenarios admit procedural skills and how the shared procedural structure should be represented across successful traces. We study this problem in FSM-defined scenarios, where successful traces can be viewed as paths in an unknown transition graph, and formulate procedural skills as reusable parameterized control-flow subgraphs. Based on this view, we introduce SkillDisCo, a distillation-and-compilation framework that distills reusable PFSM subgraphs from successful traces and compiles them into callable, executable, and verifiable procedural skills. Experiments on ALFWorld and WebArena show that SkillDisCo improves success rates and reduces agent turns across benchmarks and model scales, demonstrating the benefits of representing shared experience as reusable execution structures.
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。