Security audit
Security checks for vulnerabilities and agentic risk
This appears to be a small local regex tester, but the skill documentation and declared tool authority overstate unrelated file-changing behavior.
Review before installing. The code itself looks like a local regex testing utility, but the package should be corrected so its permissions and documentation match the actual tool, especially removing or justifying Write access and the unsupported --apply/file-management claims.
The file presents the skill guidance entirely in Chinese, including operational instructions for agents and maintainers, with no indication that language selection is optional or that alternative locales are supported. This can violate language/locale policy when a skill imposes a specific language by default without documented user opt-in or justification.
The file includes substantial licensing and governance text only in Chinese, including instructions such as '务必并读' and review requirements, without providing an equivalent English version for those key terms. This creates a language policy concern because users who rely on English are not given an explicit language choice or opt-in for critical information.
The phrase "Whenever the user needs regex-sandbox capabilities" does not define concrete activation conditions or boundaries, so it is unclear when this skill should or should not be invoked. This kind of broad wording can overlap with many ordinary requests involving text processing or debugging.
The English documentation describes disk-writing, backups, deletion, and an --apply mode that are inconsistent with the manifest, which declares a regex sandbox with filesystem: false and subprocess: false. This kind of capability mismatch is dangerous because operators and downstream agents may trust the prose over the manifest and invoke the skill in riskier contexts than intended, leading to unauthorized file modification or incorrect trust decisions.
The phrase "当用户需要正则沙盒相关能力时" is circular and does not specify what requests should activate the skill versus not activate it. Without concrete scope or examples, the activation condition is too vague for safe routing.
The Chinese section repeats claims about write-to-disk behavior, backups, moving/renaming, and duplicate deletion that contradict the stated regex-only purpose and the manifest permissions. Repeating the mismatch in another language increases the chance that a reviewer or agent will misunderstand the skill's real authority and use it under false assumptions.
The top-level description is written only in Chinese even though the file also contains English fields later, which may impose a language preference without explicit user choice. This can conflict with language/locale expectations if the user has not opted into Chinese-first behavior.
The visible natural-language content is entirely in Chinese, including the core description and file guidance, with no indication that other languages are supported. This can violate a language/locale policy when users or agents are not given an opt-in or alternative language path.
The argument parser description and help strings are written in Chinese, and later runtime messages are also Chinese-only. This imposes a specific language on all users without any opt-in or alternative locale, which matches the language/locale policy concern for natural-language content.
