| title | COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill Optimization |
|---|---|
| authors | Pingchen Lu, Xiangyi Wang, Xiang Li, Jie Mao, Zikun Qu, Junfeng Luo, Yao Shu, Bryan Kian Hsiang Low, Zhongxiang Dai |
| date | 2026-09-10 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2609.11682 |
| affiliation | Chinese University of Hong Kong, Shenzhen |
| upvotes | 43 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Large language model (LLM) agents can benefit from reusable skills distilled from prior task experience, yet existing skill optimization methods often rely on costly execution-based evaluation and substantial task data. We introduce COBRA-Skills, an efficient framework that formulates skill optimization as budgeted sequential optimization over a dynamically evolving candidate space. COBRA-Skills couples contextual-bandit-guided prioritization with evidence-grounded skill evolution, selectively allocating evaluations to promising or informative candidates while continually refining the skill population from execution feedback. Across six heterogeneous agent benchmarks and three target models, COBRA-Skills consistently achieves the strongest average performance among compared methods, while reducing optimization cost by 55--58\% relative to SkillOpt and using only 50 unique optimization examples per benchmark. Further analyses show that COBRA-Skills remains robust to changes in the agent harness and performs effectively when the target model itself is used for skill generation and refinement. |
cite
@misc{lu2026cobra,
title={COBRA-Skills: Contextual Bandit-Guided Evolution for Agent Skill Optimization},
author={Pingchen Lu and Xiangyi Wang and Xiang Li and Jie Mao and Zikun Qu and Junfeng Luo and Yao Shu and Bryan Kian Hsiang Low and Zhongxiang Dai},
year={2026},
eprint={2609.11682},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2609.11682}
}