基于跨Harness适配的检索增强技能优化
原标题：Paper page - Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation
AI 摘要
该论文提出 RASO（Retrieval-Augmented Skill Optimization）框架，利用公开技能库作为先验知识来优化 agent 技能，而非仅依赖昂贵的 agent rollout 迭代。RASO 包含两个阶段：RASI 无需 rollout 即可构建知识驱动的初始技能，RASU 则根据执行反馈检索外部知识迭代精炼技能。在四个 agent 基准和两个模型上，RASO 持续优于无检索增强的基线，并表明跨 harness 适配是关键，直接复用检索到的技能可能无效甚至有害。
正文节选
Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation Abstract An agent skill is a reusable, actionable natural-language artifact that guides an agent to perform a task effectively under a given harness. Recent studies have explored the optimization of agent skills, contributing to a growing collection of publicly available skills spanning diverse tasks, domains, and harnesses. Despite millions of publicly shared skills, existing skill optimization methods largely overlook this acc