SkillMOO: Multi-Objective Optimization of Agent Skills for Software Engineering
2026-04-10 • Software Engineering
Software EngineeringArtificial Intelligence
AI summaryⓘ
The authors created SkillMOO, a system that automatically improves collections of skills for AI coding assistants. Instead of manually adjusting these skill sets, SkillMOO uses a clever process to try out and evolve better skill combinations based on feedback from coding tasks. Their method makes the AI more successful at coding while also saving time and resources. They found that simpler, more focused skill bundles work better than large, complicated ones.
LLM-based coding agentsmulti-objective optimizationskill bundlesNSGA-IIfailure analysissoftware engineering taskspass ratecost reductionbundle pruningoptimizer agent
Authors
Jingzhi Gong, Ruizhen Gu, Zhiwei Fei, Yazhuo Cao, Lukas Twist, Alina Geiger, Shuo Han, Dominik Sobania, Federica Sarro, Jie M. Zhang
Abstract
Agent skills provide modular, task-specific guidance for LLM- based coding agents, but manually tuning skill bundles to balance success rate, cost, and runtime is expensive and fragile. We present SkillMOO, a multi-objective optimization framework that automatically evolves skill bundles using LLM-proposed edits and NSGA-II survivor selection: a solver agent evaluates candidate skill bundles on coding tasks and an optimizer agent proposes bundle edits based on failure analysis. On three SkillsBench software engineering tasks, SkillMOO improves pass rate by up to 131% while reducing cost up to 32% relative to the best baseline per task at low optimization overhead. Pattern analysis reveals pruning and substitution as primary drivers of improvement, suggesting effective bundles favor minimal, focused content over accumulated instructions.