SkillNet: Create, Evaluate, and Connect AI Skills
Abstract
Current AI agents can flexibly invoke tools and execute complex tasks, yet their long-term advancement is hindered by the lack of systematic accumulation and transfer of skills. Without a unified mechanism for skill consolidation, agents frequently ``reinvent the wheel'', rediscovering solutions in isolated contexts without leveraging prior strategies. To overcome this limitation, we introduce SkillNet, an open infrastructure designed to create, evaluate, and organize AI skills at scale. SkillNet structures skills within a unified ontology that supports creating skills from heterogeneous sources, establishing rich relational connections, and performing multi-dimensional evaluation across Safety, Completeness, Executability, Maintainability, and Cost-awareness. Our infrastructure integrates a repository of over 200,000 skills, an interactive platform, and a versatile Python toolkit. Experimental evaluations on ALFWorld, WebShop, and ScienceWorld demonstrate that SkillNet significantly enhances agent performance, improving average rewards by 40% and reducing execution steps by 30% across multiple backbone models. By formalizing skills as evolving, composable assets, SkillNet provides a robust foundation for agents to move from transient experience to durable mastery.
Cite
@article{arxiv.2603.04448,
title = {SkillNet: Create, Evaluate, and Connect AI Skills},
author = {Yuan Liang and Ruobin Zhong and Haoming Xu and Chen Jiang and Yi Zhong and Runnan Fang and Jia-Chen Gu and Shumin Deng and Yunzhi Yao and Mengru Wang and Shuofei Qiao and Xin Xu and Tongtong Wu and Kun Wang and Yang Liu and Zhen Bi and Jungang Lou and Yuchen Eleanor Jiang and Hangcheng Zhu and Gang Yu and Haiwen Hong and Longtao Huang and Hui Xue and Chenxi Wang and Yijun Wang and Zifei Shan and Xi Chen and Zhaopeng Tu and Feiyu Xiong and Xin Xie and Peng Zhang and Zhengke Gui and Lei Liang and Jun Zhou and Chiyu Wu and Jin Shang and Yu Gong and Junyu Lin and Changliang Xu and Hongjie Deng and Wen Zhang and Keyan Ding and Qiang Zhang and Fei Huang and Ningyu Zhang and Jeff Z. Pan and Guilin Qi and Haofen Wang and Huajun Chen},
journal= {arXiv preprint arXiv:2603.04448},
year = {2026}
}
Comments
http://skillnet.openkg.cn/