Coding agents are increasingly used as general-purpose problem solvers, but their flexibility does not by itself confer the domain expertise needed for specialized tasks. Recent work addresses this through \textit{agent skills}: reusable workflows, and code, that augment agents with domain-specific capabilities. Most skills today are hand-crafted, and existing evolutionary approaches optimize low-level artifacts (e.g. prompts \& code) that are tightly coupled to specific models and tasks. We introduce \textbf{EvoSkill}, a self-evolving framework that automatically discovers and refines agent skills through iterative failure analysis. EvoSkill analyzes execution failures, proposes new skills or edits to existing ones, and materializes them into structured, reusable skill folders. A Pareto frontier of agent programs governs selection, retaining only skills that improve held-out validation performance while the underlying model remains frozen. We evaluate EvoSkill on two benchmarks: OfficeQA, a grounded reasoning benchmark over U.S.\ Treasury data, where it improves exact-match accuracy by \textbf{7.3\%} (60.6\% $\to$ 67.9\%); and SealQA, a search-augmented QA benchmark with noisy retrieval, where it yields a \textbf{12.1\%} gain (26.6\% $\to$ 38.7\%). We also investigate the zero-shot transfer capabilties of skills evolved on one task to the other; in particular: skills evolved from SealQA transfers zero-shot to BrowseComp, improving accuracy by \textbf{5.3\%} without modification demonstrating that skill-level optimization produces transferable capabilities beyond the training task.
Automating Skill Acquisition through Large-Scale Mining of Open-Source Agentic Repositories: A Framework for Multi-Agent Procedural Knowledge Extraction
SkillSmith: Co-Evolving Skills and Tools for Self-Improving Agent Systems
Shanghai Jiao Tong University(上海交通大学);Eastern Institute of Technology(东部技术研究所);University of Science and Technology of China(中国科学技术大学);Southeast University(东南大学);Ningbo Institute of Digital Twin(宁波数字孪生研究所)
Automated Procedural Analysis via Video-Language Models for AI-assisted Nursing Skills Assessment
Weldon School of Biomedical Engineering, Purdue University(普渡大学生物医学工程学院);Department of Industrial and Operations Engineering, University of Michigan(密歇根大学工业与运作工程系);Edward P. Fitts Department of Industrial and Systems Engineering, North Carolina State University(北卡罗来纳州立大学工业与系统工程系);School of Nursing, Purdue University(普渡大学护理学院)
Automated discovery of symbolic laws governing skill acquisition from naturally occurring data
National Engineering Research Center of Educational Big Data, Central China Normal University(华中师范大学教育大数据国家工程研究中心);Faculty of Artificial Intelligence in Education, Central China Normal University(华中师范大学人工智能教育学部);National Engineering Research Center for E-learning, Central China Normal University(华中师范大学国家数字化学习工程技术研究中心)
Explainable AI-Powered Framework for Video-Based Skill Assessment in Cataract Surgery
K.N. Toosi University of Technology(霍加托西大学);Faculties of Electrical and Computer Engineering(电气与计算机工程学院);Applied Robotics and AI Solutions (ARAS)(应用机器人与人工智能解决方案(ARAS))
SkillAudit: From Fixed-Suite Benchmarking to Skill-Centered Assessment
Northeastern University(东北大学);City University of Hong Kong(香港城市大学);Northwestern University(西北大学);Fudan University(复旦大学);University of Science and Technology of China(中国科学技术大学);Santa Clara University(圣克拉拉大学);Fenz AI;Ohio State University(俄亥俄州立大学);University of Glasgow(格拉斯哥大学);National University of Singapore(新加坡国立大学);DeciLix Lab(DeciLix实验室)
An Ontology-Based Approach to Optimizing Geometry Problem Sets for Skill Development
Harvard University(哈佛大学);Forome Association(Forome协会);IDEXX Laboratories(爱德士实验室);Russian School of Mathematics(俄罗斯数学学校);Lexington High School(列克星敦高中);Park Hill South High School(帕克希尔南高中);Brookline High School(布鲁克莱恩高中)
What Keeps Agent Skills from Being Reusable? Evidence from 138K SKILL.md Files
The Graduate Center, City University of New York(纽约城市大学研究生中心);The Ohio State University(俄亥俄州立大学);Hunter College, City University of New York(纽约城市大学亨特学院)
SkillDRE: Dual-Stage Red-Team Evolution of Agent Skills via Pre-Execution and Runtime Feedback
Beijing University of Posts and Telecommunications(北京邮电大学);North China Electric Power University(华北电力大学);Chongqing University of Posts and Telecommunications(重庆邮电大学)