AI 智能体学会从失败中迭代，不再每次从头来
现在的 AI 智能体（比如能上网查资料、操作工具的那种）每次执行任务后，只会记住最终结果，而把中间怎么决策、怎么试错的过程全扔掉。下次遇到类似任务，它又得重新摸索一遍。这篇论文做了一个叫 SkillHone 的系统，让智能体把每次决策的完整历史——包括诊断、修改、测试结果——都存下来，下次可以直接参考之前的经验，不用再重复踩坑。在 GAIA 和 WebWalkerQA-EN 两个深度研究基准上，它比商业级的 deep-research 智能体分别高出 15.8 和 3.2 个百分点。在内部工具分析场景中，平均准确率提升 18.8 个百分点。它不是你明天就能用的产品，但指向一个方向：AI 智能体正在从“一次性执行”走向“持续学习”。
📄 原文摘要(英文)
Agent skills extend language-model agents with task-specific procedures, scripts, and references, but the tasks and environments they target continually change. Existing methods improve skills in bounded runs and retain only the final artifact, discarding the decision history that later agents need to interpret prior revisions, evaluations, and rejected alternatives. We introduce SkillHone, a harness for continual agent skill evolution grounded in persistent decision history. SkillHone pairs skill revisions with evaluation-side evidence that supplies practice feedback, recording structured histories of diagnoses, revisions, evidence, and outcomes. Role-separated subagents run candidate skills on practice probes with redacted reporting and propose revisions informed by prior decisions, enabling cross-session refinement without rediscovering past rationale. On deep-research benchmarks, SkillHone runs without a pre-integrated search stack and outperforms the commercially backed deep-research agent by 15.8 points on GAIA and 3.2 points on WebWalkerQA-EN, while also exceeding prior skill-evolution methods. We further deploy SkillHone on internal tool-mediated analysis scenarios, where it improves accuracy by an average of 18.8 points across seven settings.