From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills
2026-05-22 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors study how language agents learn and reuse 'skills', which are sets of procedures based on past experiences. They focus on two types of skills: those related to specific domains and those generated by models themselves. By testing these skills across different tasks, they find that while model-generated skills often help, they can sometimes harm performance, and success depends on both how skills are created and used. The authors analyze each stage of skill development and introduce a new method called a 'meta-skill' to improve skill quality and reduce negative effects across various domains.
language agentsskillsdomain-level skillsmodel-generated skillsskill extractionskill consumptionnegative transferexperience generationmeta-skilltask domains
Authors
Zisu Huang, Jingwen Xu, Yifan Yang, Ziyang Gong, Qihao Yang, Muzhao Tian, Xiaohua Wang, Changze Lv, Xuemei Gao, Qi Dai, Bei Liu, Kai Qiu, Xue Yang, Dongdong Chen, Xiaoqing Zheng, Chong Luo
Abstract
Language agents increasingly improve by reusing \emph{skills} -- structured procedural artifacts distilled from past experience. In particular, \emph{domain-level} and \emph{model-generated} skills are especially promising. They offer fast adaptation within a domain by encoding domain-specific recurring procedures, and they scale beyond labor-intensive hand-crafting. However, while extraction methods continue to proliferate, understanding remains limited, with no comprehensive study spanning the full skill lifecycle -- \textbf{experience generation}, \textbf{skill extraction}, and \textbf{skill consumption} -- to ask whether such skills actually work, when they work, and what makes them succeed or fail. To close this gap, we build a utility-grounded evaluation framework that provides systematic experimental results across extractors and target agents, covering five diverse agentic task domains. We find that model-generated skills are beneficial on average but exhibit non-trivial negative transfer, and that neither extractors nor targets behave uniformly. A model can be a strong extractor yet a weak consumer, or vice versa, with skill utility independent of model scale or baseline task strength. To explain these patterns, we then dissect each lifecycle stage in depth, analyzing how experience composition shapes skill quality, what properties characterize useful skills, and how the same skill transfers across different consumers. Finally, we translate these findings into a concrete \emph{meta-skill} that guides skill extraction toward the features tied to actual utility, which consistently improves skill quality across domains and substantially reduces negative transfer.