Security audit
Security checks for vulnerabilities and agentic risk
This is a Markdown-only pet behavior planning skill with no code execution, credential access, persistence, or external data transfer.
Reasonable to install if you want pet behavior plan drafting help. Treat outputs as general guidance, avoid sharing unnecessary personal details, and use a veterinarian or qualified behavior professional for aggression, self-harm, severe anxiety, medical symptoms, or safety-critical cases. The publisher should clarify which features are free versus Pro.
The trigger keywords include broad, common phrases such as '拆家' or '宠物训练' that can appear in ordinary conversation without clear intent to invoke this skill. Overly broad activation can cause the wrong skill to engage, leading to unintended processing of user-provided personal details or confusing workflow hijacking in multi-skill environments.
L20 明确写明“永久免费，无需付费、无需授权（MIT-0）”，给人的含义是该技能能力范围内不存在付费限制。但 L69 又写“批量处理、进阶段输出（异议应对/执行排期/记录表）与行业模板库属于增强能力，见同名 Pro 版本”，其中“记录表”还与前文 L26/L74 承诺的基础输出内容重叠，构成文档内部对能力边界与付费属性的直接冲突。
L26 将“进度记录表”列为该技能的核心产出之一，L74 的示例输出也包含“记录表”，表明这是默认应提供的能力。但 L69 又称“进阶段输出（异议应对/执行排期/记录表）”属于增强能力并指向 Pro 版本，导致文档对同一输出是否属于基础免费能力产生直接矛盾。
The natural-language examples include generic follow-up instructions like '按上面的结果改成更简短的版本' and '把这些整理成一张表', which are not uniquely scoped to this skill. In a shared assistant environment, this ambiguity can cause accidental routing, context bleed, or unintended continuation of the skill when the user may be addressing another tool or asking for a general transformation.
No suspicious patterns detected.
