测试智能体能否从自身运行中提炼可复用技能，提升后续任务表现。
EvoClawBench: Can Agents Learn Reusable Skills from Their Own Runs?
- 设计闭环学习评估框架，对比无技能、预设技能和后置总结技能三种模式。
- 实验显示部分模型技能提升显著，但多数模型性能反而下降，存在严重不稳定性。
- 适合关注智能体自我进化能力与技能泛化性的研究者参考。
现有智能体评测侧重任务完成、工具使用或技能效用，却未检验运行过程能否将自身经验转化为可复用的技能以提升后续执行。我们提出EvoClawBench，一个针对重复性、基于固定任务的闭环技能学习评测基准。该基准比较直接执行、执行前预设技能（PreSkill）以及首次运行后基于证据总结技能（PostSkill）并进行第二次执行的效果。基准包含100个任务和502个子问题，涵盖编码、数据、办公、安全、运维及领域文档工作流，支持多种智能体运行时。在本地执行下对OpenClaw和nanobot的实验表明，直接基线性能强烈依赖运行时：OpenClaw在所有模型下均低于20%，而nanobot则在56.45%至96.13%之间波动。自生成技能效果参差不齐：nanobot GPT-5.4在各模式下均保持96%以上，MiniMax-M2.7从90.97%提升至94.50%；但nanobot DeepSeek-V4-Pro在PreSkill下由77.77%降至4.80%，PostSkill下更跌至0.99%。OpenClaw也表现出非单调行为，部分技能运行接近基线，部分严重崩溃。结果表明，从自身运行中学习可复用技能具有选择性和成本敏感性，并非自动带来性能提升。
原文摘要 · Abstract (English)
Existing agent benchmarks primarily test task completion, tool use, or skill utility, but do not isolate whether a runtime can convert evidence from its own runs into reusable skills that improve fresh executions after authoring overhead. We introduce EvoClawBench, a benchmark for this closed-loop skill-learning question on repeated, fixture-backed tasks. EvoClawBench compares direct execution without skills, PreSkill authoring before execution, and PostSkill summarization from first-run evidence followed by a fresh second execution. The suite contains 100 tasks and 502 sub-problems across coding, data, office, security, operations, and domain-document workflows, with support for multiple agent runtimes. Experiments with OpenClaw and nanobot under local execution show that direct baseline performance is strongly runtime-dependent: OpenClaw remains below 20% across models, while nanobot ranges from 56.45% to 96.13%. Self-authored skills have mixed effects. nanobot GPT-5.4 stays above 96% in all modes and MiniMax-M2.7 improves from 90.97% to 94.50% under PostSkill, but nanobot DeepSeek-V4-Pro drops from 77.77% to 4.80% with PreSkill and 0.99% with PostSkill. OpenClaw shows similarly non-monotonic behavior, with some skill runs near baseline and others collapsing. These results indicate that learning reusable skills from an agent's own runs is selective and cost-sensitive, rather than an automatic benefit of adding skill authoring to an agent loop.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。