📄 中文摘要
当前的人工智能代理能够灵活调用工具并执行复杂任务，但其长期发展受到技能系统性积累和转移缺乏的制约。缺乏统一的技能整合机制，导致代理经常在孤立的上下文中“重新发明轮子”，而未能利用先前的策略。为了解决这一限制，提出了SkillNet，一个旨在大规模创建、评估和组织人工智能技能的开放基础设施。SkillNet在统一的本体结构中组织技能，支持从异构来源创建技能，建立丰富的关系连接，并在安全性、完整性、可执行性和可维护性等多个维度上进行评估。
SkillNet：创建、评估和连接人工智能技能
出处: SkillNet: Create, Evaluate, and Connect AI Skills
发布: 2026年3月6日
📄 English Summary
SkillNet: Create, Evaluate, and Connect AI Skills
Current AI agents are capable of flexibly invoking tools and executing complex tasks; however, their long-term advancement is hindered by the lack of systematic accumulation and transfer of skills. Without a unified mechanism for skill consolidation, agents often 'reinvent the wheel', rediscovering solutions in isolated contexts without leveraging prior strategies. To address this limitation, SkillNet is introduced as an open infrastructure designed to create, evaluate, and organize AI skills at scale. SkillNet structures skills within a unified ontology that supports the creation of skills from heterogeneous sources, establishes rich relational connections, and performs multi-dimensional evaluations across Safety, Completeness, Executability, and Maintainability.