今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution | arXiv | 2026-08-27 | AI-Agent | 高 |
| 2 | Do User-Authored Permission Policies Improve Protection Against AI Agent Overreach? | arXiv | 2026-08-27 | AI-Agent | 高 |
重点论文与技术动态
1. WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution
- 来源：arXiv
- 日期：2026-08-27
- 作者/机构：Liyan Tang, Cyrus Rashtchian, Chun-Sung Ferng, Andrew Tomkins, Da-Cheng Juan, Tu Vu
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.AI, cs.CL
一句话结论
WikiSkill通过持久知识库系统积累agent经验，显著提升技能进化效果。
该框架在多样基准测试中验证了系统性积累agent经验对开发可重用和可迁移技能的益处，同时ablation studies确认wiki知识积累对技能进化至关重要。
核心内容
- WikiSkill分离原始执行经验、累积知识和可执行技能，并持续将经验整合到持久知识库（wiki）中，供后续技能更新构建使用。
- 在多种基准测试和模型中，WikiSkill consistently outperforms state-of-the-art技能进化方法，优于无技能基线。
- 技能进化与模型缩放互补：大模型受益更多，小模型使用技能可显著超越大模型无技能。
方法与数据
- 摘要未明确具体方法细节。
- 摘要提到在多样基准测试和模型中验证。
价值判断
- 值得关注：技能进化与模型缩放互补，小模型有技能可超越大模型无技能，证明技能对小模型提升效果显著。
- 可复用点：持久知识库机制可推广至其他AI agent系统，实现技能的系统性积累和跨模型迁移。
- 局限/待核查：摘要未明确提及局限。
摘要
Agent skills package specialized knowledge and workflows into reusable resources that extend AI agent capabilities. Recent work automatically discovers such skills from agent experience, which enables agents to progressively adapt through interaction. However, the insights that guide skill development typically remain scattered across optimization histories, limiting their systematic reuse across iterations. We introduce WikiSkill, a framework that co-evolves agent skills with a persistent knowledge base (wiki). At a high level, WikiSkill separates raw execution experience, accumulated knowledge, and executable skills, while continuously consolidating experience into the wiki, which subsequent skill updates can build on. Across diverse benchmarks and models, WikiSkill consistently outperforms state-of-the-art skill-evolution methods and improves over no-skill baselines in most model-benchmark settings. We find that skill evolution complements model scaling: larger models generally benefit more from evolved skills, while smaller models with skills can outperform substantially larger models without them. We also find that evolved skills transfer effectively across models and model families, and skills evolved by other models can outperform self-evolved skills. Finally, our ablation studies confirm that persistent knowledge accumulation in the wiki is critical for effective skill evolution. These results demonstrate the benefits of systematically accumulating and refining agent experience for developing reusable and transferable skills.2. Do User-Authored Permission Policies Improve Protection Against AI Agent Overreach?
- 来源：arXiv
- 日期：2026-08-27
- 作者/机构：Ting Yan
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.HC, cs.CR
一句话结论
用户自定义权限策略未能有效提升对AI代理越权行为的防护，反而在阻止越权方面效果较差。
核心内容
- POLICY策略比HITL和AUTO策略更少阻止越权行为（分别低20.1%和14.5个百分点）。
- POLICY降低运行时提示数量（从18.0降至10.9），但总干预时间未显著减少（含规则设置时间）。
- 用户多选择“ask”规则（114/140），导致多数越权行为需人工审批，而非自动处理。
方法与数据
- 113名无专业软件背景参与者在HITL、AUTO、POLICY三种条件下测试，模拟18个动作（含7个越权动作）。
- 摘要未明确具体规则设置细节。
价值判断
- 值得关注：用户频繁选择“ask”揭示偏好与承诺的差距，保留逐案决策但未提前固定政策。
- 可复用点：语言模型映射动作至平实语言后果类别，提供可理解的规则机制。
- 局限/待核查：用户自定义规则未增强防护，许多越权行为在用户批准后仍执行。