This work proposes SkillLift, which decouples skill search from oracle cost by learning an oracle-aligned rubric as a structured evaluation space, and outperforms existing auto-skill methods with 40--70\% less token cost compared to frontier evolving methods.
Abstract
LLM-based agents increasingly rely on persistent skills, i.e., reusable procedural prompts, to adapt without weight updates. Existing skill self-evolution methods directly revise skill text based on execution feedback, but each oracle evaluation requires a full agent rollout, creating a supervision bottleneck that confines search to failure-patching updates. Our key insight is that ranking is a smoother supervision target than absolute outcome regression: identifying which skill is better requires fewer oracle evaluations than predicting exact scores. Building on this insight, we propose SkillLift, which decouples skill search from oracle cost by learning an oracle-aligned rubric as a structured evaluation space. We formalize this as a bilevel optimization problem solved via alternating optimization: an inner loop uses the frozen rubric as a cheap surrogate to guide skill revision at no oracle cost, while an outer loop invokes a small number of oracle rollouts to re-align the rubric via rank correlation, amortizing oracle cost and stabilizing text-space updates. Experiments on complex agent task benchmarks show that our method outperforms existing auto-skill methods with 40--70\% less token cost compared to frontier evolving methods. Codes are available at https://github.com/WalteR-MittY-pro/SkillLift.
SkillFM (Skill Flow Matching), a generative framework that synthesizes task-conditioned textual skills directly without test-time skill retrieval, is introduced, demonstrating that latent skill generation is an effective alternative to retrieval-based skill augmentation.
Zu-Ming Zhang, Jie He, Yi-Zhe Zhang et al.· arXiv.org· 0 citations
Anthropic's Agent Skills package reusable procedural know-how for an LLM agent into SKILL.md directories, and open-source aggregations have grown past 230,000 skills, making selection rather than authoring the bottleneck. The standing answer in the literature outsources selection to the agent itself: an LLM-mediated re...
Guanqun Yang, Wen-Long Zhang, Tian Shi et al.· arXiv.org· 3 citations
Multi-Objective In-context Knowledge Editing (MO-IKE), a multi-objective RL algorithm that formulates prompt construction for in-context knowledge editing as a Constrained Markov Decision Process, enabling more balanced and globally coherent prompt construction.
Xu-Zhong Wang, Maiqi Jiang, Tejal Nair et al.· 1 citation
SkillEvoReg, a general regularization framework for skill evolution inspired by anti-overfitting techniques in neural-network training, is introduced and suggested that explicit regularization is a useful complement to increasingly capable skill updaters.
Guan-Yu Nie, Fang-Zhou Zhu, Shi-Xiong Kai et al.· arXiv.org· 3 citations
Across several reasoning tasks and model families, the proposed frontier learning approach consistently achieves higher relative gains over fixed-pool baselines, demonstrating that effective post-training requires not only selecting useful problems, but continually generating them at the edge of capability.
Robin Faro, S. Ramesh, Ilija Bogunovic et al.· arXiv.org· 0 citations
In-context learning (ICL) is crucial for boosting the inference performance of large language models (LLMs). However, the effectiveness of ICL in LLMs is greatly influenced by the choice of demonstration sets. Exhaustive searches over these sets are combinatorial, and existing selectors often rely on relevance or likel...