The arXiv abstract for SkillNEO frames agent skills as reusable artifacts that let LLM agents accumulate and refine expertise through experience. Building on this idea, the authors describe skill evolution as an iterative process that improves skills using execution feedback.
SkillNEO is presented as a method for this evolution, with the title specifying a "non-elitist" evolutionary optimization approach. The abstract itself does not elaborate on how the non-elitist strategy works or how it compares with elitist alternatives, and no experimental results are included in the announcement.
This line of research points toward agents that can steadily build up procedural knowledge over time, rather than relying solely on static prompts or fixed toolkits. The brief abstract leaves many specifics open, but the core motivation is clear: turn experience into lasting, improvable skills.