Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed local knowledge-base organizer, but users should treat its deletion and rename examples carefully because they can change or remove files.
Install only if you are comfortable with a Chinese-language skill that guides local file organization. Before allowing it to rename or delete anything, run dry-run or review steps first, make backups, and explicitly confirm the exact files to be changed or removed.
声明将该技能描述为可整理本地知识库，并支持文件合并、重命名、去重、索引更新等完整整理流程，还声称适用于多种文件类型。但代码实际行为较有限：它只扫描 .md 文件，打印文件列表和标题，检查命名格式；所谓“执行整理”部分并没有具体重命名、README 更新、合并或去重实现，仅有占位注释和成功提示。因此，代码的真实能力显著弱于声明，且适用范围也更窄。虽然它确实与知识库整理主题相关，但描述夸大了已实现功能，构成描述与行为不匹配。
The manifest description says the skill should be used whenever the user needs to organize a knowledge base, merge duplicate content, unify file naming, or update indexes, and it also applies to Markdown, scripts, and config files. This is a wide natural-language trigger without clear boundaries, exclusions, or specific invocation phrases, which could cause unintended activation for general document-organization requests.
The skill’s manifest description and operational guidance are entirely in Chinese, including expected usage conditions and examples, with no indication that the user may choose another language. This creates a language/locale constraint without opt-in or justification as a region-specific tool.
The skill includes direct destructive deletion commands (rm) immediately after merge guidance without an explicit confirmation checkpoint at the point of deletion. In a file-management skill operating on a local knowledge base, this materially increases the risk of accidental or premature data loss, especially if merges were incomplete, backups were not actually created, or file selection was wrong.
This shell script includes its title, usage instructions, help output, and prompts exclusively in Chinese. Under the stated policy, forcing a specific language without offering the user a choice is a natural-language policy violation.
No suspicious patterns detected.
