SkillBrew: Multi-Objective Curation of Skill Banks for LLM Agents
For developers of retrieval-augmented LLM agents, this work addresses the practical problem of skill bank bloat by introducing a principled curation method, though the gains are incremental over existing append-only approaches.
SkillBrew formulates skill bank curation for LLM agents as a constrained multi-objective optimization problem, balancing utility, diversity, and coverage. It outperforms append-only baselines on two public benchmarks, demonstrating that principled curation improves agent performance.
Retrieval-augmented LLM agents increasingly rely on curated skill banks: collections of reusable textual principles that guide decision making on complex tasks. Existing approaches typically expand these banks in an append-only fashion, continuously adding new skills without removing redundant, outdated, or harmful ones, resulting in inefficient and poorly curated repositories. In this paper, we formulate the skill bank curation as a constrained multi-objective problem: a desirable bank must be useful for the agent, diverse in its content, and provide good coverage of the query distribution. To this end, we introduce SkillBrew, a multi-objective curation framework that formalizes skill bank curation as Pareto-aware optimization under a utility constraint, and solves it via a bi-level propose-then-verify loop. We evaluate our approach on two public benchmarks. Our findings suggest that treating skill banks as objects of principled curation, rather than ever-growing append-only logs, is an important step toward building self-improving LLM agents.