Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed Chinese-language workflow for creating and reviewing Agent Skills, with local helper scripts that match that purpose.
Before installing, expect this skill to guide an agent through creating or changing skill files and optionally running local Node.js helper scripts. Review generated skills before publishing, keep the explicit confirmation gates for destructive or external actions, and note that most bundled guidance is written in Chinese.
声明描述的是一个面向技能全生命周期的生产级能力集合，包括创建、重构、安全评审与评估/发布流程建设；而提供的代码只是一个单独的 Node.js 评分工具，用于检查某个 skill 目录是否包含特定文件结构和文档要素，并据此计算分数与等级。虽然“评估”与声明中的一部分有轻微相关性，但代码并未实现从0到1创建技能、重构技能、建立发布流程，安全评审也仅限于检查 SKILL.md 是否包含安全相关关键词，属于非常有限的静态检查。因此其主要目的与声明相比明显更窄且不同，构成描述与行为不匹配。
Referenced artifact was not completely inspected
- `SKILL.md`
Referenced artifact was not completely inspected
- `node scripts/score-skill.js <skill-dir>`
Referenced artifact was not completely inspected
- `node scripts/score-skill.js <skill-dir>`
The README suggests activating the skill with very broad, everyday phrases like '帮我做一个 skill' and '对这个 skill 做安全审查'. In systems that auto-activate skills based on user phrasing, generic triggers can cause unintended invocation in unrelated conversations, leading to prompt/context injection into workflows the user did not explicitly request.
The file’s description and operating instructions are written as mandatory Chinese-language behavior, but there is no opt-in mechanism or statement that the skill can adapt to the user’s preferred language. Under the language/locale policy, forcing a specific language without user choice is a natural-language policy concern.
The activation trigger includes broad phrases such as requests to 'help make a skill' or create reusable templates, which can overlap with ordinary user requests and cause unintended auto-activation. In a skill-capable agent environment, overbroad triggering can hijack unrelated conversations, alter agent behavior unexpectedly, and expand the influence of this skill beyond intended contexts.
This markdown file presents the entire release checklist in Chinese and does not indicate that language selection is optional or that the checklist is intended only for a Chinese-language team or region-specific workflow. Under the policy rule, forcing a specific language without opt-in is a natural-language policy concern.
The template leaves the trigger condition as a free-form placeholder without requiring scope boundaries, disallowed triggers, or precision criteria. In a skill-forging context, this can propagate into skills that activate on vague or overly broad phrases, increasing the chance of unintended execution, context confusion, or invocation during unsafe workflows.
This markdown file presents all skill instructions in Chinese and uses mandatory language such as '必须' and '适用于' throughout, with no indication that users may choose another language or locale. Under the policy criteria, forcing a specific language without user opt-in is a natural-language policy violation.
SQP-3 applies to all file types and covers natural-language policy violations such as forcing a specific language without user opt-in. This reference document presents all instructions in Chinese and does not indicate that the skill is China-specific or that users may choose another language.
The title and the entire document are written in Chinese, presenting the standard as a universal 'cross-platform' skill standard without any indication that Chinese is optional or region-specific. Under the policy, forcing a specific language without user opt-in or documented justification is a natural-language locale policy violation.
The document title and instructional content are written in Chinese, which imposes a specific language on the user without offering a language choice or stating that the skill is region-specific. The policy for this audit flags language or locale constraints unless the skill explicitly provides opt-in or justification.
This code emits user-facing status messages exclusively in Chinese (for example, the check labels in the scoring report), while the CLI usage and error strings elsewhere are in English. That creates an inconsistent and effectively forced locale experience without any language choice or justification, which matches the language/locale policy violation criteria.
The natural-language user-facing description and usage instructions are written in Chinese, but the file does not indicate that this is a China-specific skill or offer users a language/locale option. Under the stated policy, forcing a specific language without opt-in is a potential locale-policy violation.
This markdown file contains user-facing instructions primarily in Chinese, but it does not state that the exercise is intended only for Chinese-speaking users or provide an opt-in language choice. Under the policy, forcing a specific language without user choice can be a natural-language policy violation.
No suspicious patterns detected.