Skill VM: Write Once, Run Everywhere Efficiently
Le Chen, Erhu Feng, Yubin Xia, Haibo Chen
2026年份
摘要
LLM agents increasingly adopt skills as a reusable unit of composition. While skills are shared across diverse agent platforms, current systems treat them as raw context, causing the same skill to behave inconsistently for different agents. This fragility undermines skill portability and execution efficiency.
问问这篇 Paper
问问你的智能体。
Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。
相关 Paper
- PragLocker: Protecting Agent Intellectual Property in Untrusted Deployments via Non-Portable PromptsQinfeng Li, Yuntai Bao, Jianghui Hu, Wenqi Zhang 等ICML 2026
- PolySkill: Learning Generalizable Skills Through Polymorphic Abstraction For Continual LearningSimon Yu, Gang Li, Weiyan Shi, Peng QiICLR 2026 · 被引用 11 次
- Efficient Skill Grounding via Code Refactoring with Small Language ModelsSera Choi, Wonje Choi, Saehun Chun, Daehee Lee 等ICML 2026
- SkillScope: Toward Fine-Grained Least-Privilege Enforcement for Agent SkillsJiangrong Wu, Yuhong Nan, Yixi Lin, Huaijin Wang 等CCS 2026 · 被引用 5 次
- "Do Not Mention This to the User": Detecting and Understanding Malicious Agent Skills in the WildYi Liu, Zhihao Chen, Yanjun Zhang, Gelei Deng 等USENIX Security 2026 · 被引用 46 次
