研究：相关技能也可能让 Agent 更差
Very timely paper. An agent skill can be completely relevant to the task and sti…
Very timely paper. An agent skill can be completely relevant to the task and still make the agent worse.
这篇论文非常及时。一个智能体技能即使与任务完全相关，也可能让智能体表现更差。
This study compares the same tasks under different skill setups while keeping the model, agent framework, repository, and verifier fixed.
本研究在保持模型、智能体框架、代码库和验证器不变的情况下，比较了不同技能配置下的相同任务。
Across 307 confirmed skill-induced failures, the bigger problem was not obviously irrelevant skills. Among 125 functional failures, 86 came from task-implementation faults: the skill pushed the agent to fill a required element incorrectly or omit it entirely.
在307个经确认的技能引发的失败中，更大的问题并非明显不相关的技能。在125个功能性失败中，86个源于任务实现缺陷：技能促使智能体错误地填写了必需元素或完全遗漏了它们。
Cost failures had a similar pattern. Among 182 high-confidence efficiency regressions, 114 came from extra procedure, not just longer prompts. Excessive verification alone caused 67 cases, with skills turning tests, rebuilds, debugging, and checklists into mandatory work.
成本失败也有类似模式。在182个高置信度的效率退化案例中，114个源于额外程序，而不仅仅是更长的提示词。仅过度验证就导致了67个案例，技能将测试、重建、调试和检查清单变成了强制性工作。
The practical warning is simple: a skill can be perfectly on-topic and still make an agent worse.
实际警示很简单：一个技能可以完全切题，但仍然会让智能体变得更差。
So for agent platforms, loading a skill should be treated like changing system behavior: check it against task requirements, compare it with a no-skill run, and measure the extra actions it induces.
因此，对于智能体平台而言，加载技能应被视为改变系统行为：需对照任务要求进行核查，与无技能运行进行比较，并衡量其引发的额外操作。
– arxiv. org/abs/2608.11888
– arxiv. org/abs/2608.11888
Title: "Agent Skills Can Be Harmful: An Empirical Study of Skill-Induced Failures in LLM Agents"
标题：“智能体技能可能有害：LLM智能体中技能引发失败的实证研究”
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力