让大模型智能体在出错时精准定位问题步骤并修复技能
SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories
- 按步骤分析失败轨迹，精准定位首个可行动故障点
- 针对具体错误步骤更新技能，提升任务成功率1.5~1.8个百分点
- 无需训练即可维护技能，适合需要可靠推理的复杂任务场景
大型语言模型（LLM）智能体越来越依赖可复用的外部技能来完成长周期交互任务。现有免训练技能适配方法通常基于完整轨迹或会话级反馈进行更新，导致错误归因粗糙，常引发不稳定或过度宽泛的修改。我们提出 SkillAdaptor，一种免训练的细粒度步骤级技能适配框架，具备明确的错误归因能力，可集成至 OpenClaw 类智能体系统中。给定失败轨迹后，SkillAdaptor 识别首个可行动故障步骤，将责任关联至候选技能，并在显式验证条件下进行针对性更新，同时保持主干模型冻结。在 WebShop、PinchBench 与 Claw-Eval 数据集上，使用 Kimi-K2.5、GLM-5、GPT-5.2 进行评估，SkillAdaptor 在三项基准测试中均优于无技能及技能适配基线，单指标最高提升分别为：PinchBench 平均得分 +1.5%，Claw-Eval 平均得分 +1.8%，WebShop 成功率 +1.7%。结果表明，步骤级归因有助于实现更稳定、可审计的免训练技能维护。
原文摘要 · Abstract (English)
Large language model (LLM) agents increasingly rely on reusable external skills to solve long-horizon interactive tasks. Existing training-free skill adaptation pipelines usually update skills from full trajectories or session-level feedback, which makes failure attribution coarse and often produces unstable or overly broad revisions. We propose SkillAdaptor, a training-free step-level skill adaptation framework with explicit failure attribution, and it can plug into OpenClaw-class agent harnesses. Given a failed trajectory, SkillAdaptor identifies a first actionable fault step, links responsibility to candidate skills, and applies targeted updates under explicit acceptance checks while keeping the backbone frozen. We evaluate on WebShop, PinchBench, and Claw-Eval with Kimi-K2.5, GLM-5, and GPT-5.2. SkillAdaptor improves over no-skill and skill-adaptation baselines on all three suites, with the largest single-metric improvements of +1.5 points on PinchBench Avg Score%, +1.8 on Claw-Eval Avg Score, and +1.7 on WebShop success rate. These results indicate that step-level attribution supports more stable and auditable training-free skill maintenance\footnote{The code will be released at https://github.com/zjunlp/SkillAdaptor.}.
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。