SkillOpt-Lite：透過一行Vibe實現更優更快之智能體自我演化

摘要

虽然自主智能体的技能优化已日益受到关注，但现有方法依赖于复杂的流水线。这留下了一个根本问题尚未得到解答：什么构成了技能优化的最小可行流水线，其中每个组件都经过理论或经验必要性的证明？我们通过零阶优化（ZO）将技能优化形式化，将经典方法（中心差分、信任区域）与近期文献对应起来。我们注意到，与经典ZO中盲目的数值扰动不同，技能轨迹可作为可解释的调试反馈。基于Claude Code理念和PAC学习，我们建立了收敛与泛化的三条原则：基于文件系统的轨迹探索、共识属性挖掘、以及独立验证门控。去除冗余后，我们提出了SkillOpt-Lite。它加速了收敛，并全面超越了完整版SkillOpt：在GPT-5.5上提升LiveMath +8.8分，在GPT-5.4-nano上提升+25.4分，使nano模型超越了经SkillOpt优化的标准GPT-5.4。最后，我们将框架集成到VSCode Copilot等生产级编码智能体中，使开发者能够通过一条情境指令演化智能体技能。由于我们的框架将所有智能体组件视为标准可编辑代码，这一最小流水线自然可以推广到完整的工具优化（HarnessOpt）。在SpreadsheetBench上，HarnessOpt使GPT-5.4-nano达到0.7758的准确率，超越了使用标准流水线的更大模型GPT-5.5（0.7620）。代码已开源至https://github.com/EvolvingLMMs-Lab/SkillOpt-Lite。

English

While skill optimization for autonomous agents has gained traction, existing methods rely on complex pipelines. This leaves a fundamental question unaddressed: What constitutes a minimal viable pipeline for skill optimization, where every component is justified by theory or empirical necessity? We formalize skill optimization via Zeroth-Order (ZO) optimization, mapping classical counterparts (central difference, trust regions) to recent literature. Noting that unlike blind numerical perturbations in classical ZO, skill trajectories serve as interpretable debugging feedback. Grounded in Claude Code philosophy and PAC learning, we establish three principles for convergence and generalization: file-system-based trajectory exploration, consensus attribute mining, and independent validation gating. Eliminating redundancies, we propose SkillOpt-Lite. It accelerates convergence and outperforms full SkillOpt: improving LiveMath by +8.8 points on GPT-5.5 and +25.4 points on GPT-5.4-nano, allowing the nano model to surpass standard GPT-5.4 optimized by SkillOpt. Finally, we integrate our framework into production coding agents like VSCode Copilot, enabling developers to evolve agent skills via one line of vibe. Because our framework treats all agent components simply as standard editable code, this minimal pipeline naturally generalizes to full harness optimization (HarnessOpt). On SpreadsheetBench, HarnessOpt enables GPT-5.4-nano to achieve 0.7758 accuracy, outperforming the larger GPT-5.5 running standard pipelines (0.7620). Code is available at https://github.com/EvolvingLMMs-Lab/SkillOpt-Lite.