Security audit
Security checks for vulnerabilities and agentic risk
This skill coherently generates printable Chinese kindergarten activity plans and its file-writing behavior is limited to user-directed HTML output.
Install this if you want Chinese printable activity plans for kindergarten-aged children. Be aware that it may activate on broad child/activity wording and it writes an HTML file to the path chosen by the user or agent, so review the output location when running it.
The entire skill description and examples are presented only in Chinese, with no indication that users may choose another language or that the skill is intentionally restricted to a Chinese-speaking context. Under the stated policy, forcing a specific language without opt-in can be a natural-language policy violation.
The invocation guidance is broad enough to capture many ordinary requests about children, activities, games, crafts, and family plans, which can cause the agent to route unrelated or only partially related user queries into this skill. Over-broad routing is dangerous because it increases misfires, reduces user intent fidelity, and may suppress use of more appropriate skills or general reasoning paths.
The skill instructs the agent to run a script that writes output files (for example via --out plan.html) but does not declare any permissions or allowed-tools scope. That creates an authorization gap where file-writing behavior is implied by the prompt rather than explicitly constrained, increasing the risk of unintended or overbroad filesystem access if the runtime permits tool use.
The description says to use the skill whenever the user mentions very common childcare-related terms like 幼儿, 孩子, 活动, 游戏, 亲子活动, or 周计划. Such broad routing criteria can cause the skill to activate for many ordinary requests that may not actually be asking for this specific workflow, leading to skill hijacking or incorrect tool/file actions in unrelated conversations.
The trigger table explicitly treats generic words such as 健康, 语言, 社会, 科学, 艺术, 活动方案, and even English words like activities or experiment as sufficient triggers. This makes invocation highly collision-prone and can redirect broad classes of normal conversation into this skill, especially when paired with instructions to generate files and plans.
The rule "语言输出用中文" forces a specific language for all outputs. This is a natural-language policy concern because it removes user language choice and does not present the language restriction as an opt-in or clearly justified region-specific requirement.
The trigger condition for applying the weekly plan template is broad and underspecified, so the skill may activate for generic requests about a 'weekly plan' without sufficient confirmation that the user wants kindergarten activity scheduling. In an agent environment, this can cause incorrect tool/skill routing, irrelevant output, or accidental override of user intent, especially when adjacent skills handle other kinds of weekly plans or educational tasks.
This markdown file contains user-facing natural language exclusively in Chinese for the release notes. Under the policy, forcing a specific language without user opt-in can be a locale-policy violation, and there is no indication here that the skill is region-specific or that users can choose another language.
SQP-3 适用于所有文件类型。该模板从标题到指令均固定为中文，且未提供用户语言选择或 opt-in 说明；如果该技能面向通用用户，这种默认单一语言输出可能违反语言选择策略。
The output explicitly sets lang="zh" and all user-facing text is hardcoded in Chinese, which imposes a specific language experience. The file does not provide any option to select another language or explain that the tool is intentionally limited to a Chinese-only context.
