SkillZip：通过发现可重用结构实现自进化智能体的免评估技能压缩

中文摘要 AI 辅助

自进化智能体通过追加成功的过程与失败修复来积累可重用技能。随着时间推移，同一需求常被在多个分支、示例和警告中重复表述，而常见动作序列被复制而非重用，导致技能注入成本高昂且难以维护。通用提示压缩不适用于此场景，因为技能并非扁平文本：其名称和描述定义适用时机，工作流控制执行，工具与输出契约约束有效性，即使无采样任务激活，罕见异常也可能至关重要。评估引导压缩可测试这些行为，但会引入 rollout（试执行）、成本，并依赖压缩时的评估集。我们提出 SkillZip，一种免评估方法，通过寻找技能最短的忠实结构解释来压缩技能。其核心思路是“一次解释，多次引用”：在适用范围处单次陈述重复规则，将重复动作序列提炼为共享过程，仅保留差异作为显式异常。我们将此思路形式化为基于技能契约与残差的类型化最小描述长度目标，需满足每个提取的触发器、工作流边、工具需求、义务及输出字段的硬覆盖约束。该公式提供简单的共享阈值，通过构造保留唯一罕见规则，支持高效局部更新。SkillZip 具有单次模式（含一次结构化提取调用与确定性优化），以及持续的“写时压缩（Zip-on-Write）”模式，可整合每次自进化补丁，无需重放任务或重新解析完整历史。通过全面实验评估，我们证明 SkillZip 在压缩性能、可泛化性及成本开销方面的有效性与优越性。

英文摘要

Self-evolving agents accumulate reusable skills by appending successful procedures and failure fixes. Over time, the same requirement is often restated in several branches, examples, and warnings, while common action sequences are copied rather than reused. The resulting skill becomes expensive to inject and difficult to maintain. Generic prompt compression is ill-suited to this setting because a skill is not a flat passage: its name and description define when it applies, its workflow controls execution, its tool and output contracts constrain validity, and rare exceptions may remain essential even when no sampled task activates them. Evaluation-guided compression can test these behaviors, but it introduces rollouts, cost, and dependence on the compression-time evaluation set. We present SkillZip, an evaluation-free method that compresses a skill by finding its shortest faithful structural explanation. The intuition is explain once, reference many: state a repeated rule once at the scope where it applies, factor a repeated action sequence into a shared procedure, and keep only the differences as explicit exceptions. We formalize this intuition as a typed minimum description-length objective over a skill contract and a residual, subject to a hard coverage constraint for every extracted trigger, workflow edge, tool requirement, obligation, and output field. The formulation provides simple sharing thresholds, preserves unique rare rules by construction, and supports efficient local updates. SkillZip has a one-shot mode with one structured extraction call and deterministic optimization, and a continual Zip-on-Write mode that integrates each self-evolution patch without replaying tasks or reparsing the full history. Through comprehensive experimental evaluations, we demonstrate the effectiveness and superiority of SkillZip in compression performance, generalizability, and cost overhead.

发表机构

• Alibaba Group（阿里巴巴集团）
• Zhejiang University（浙江大学）
• Duke University（杜克大学）

机构由 AI 辅助整理，请以论文原文为准。