Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese-language industry best-practices reference skill with no executable code, hidden commands, data exfiltration, or persistence behavior beyond disclosed content auto-update metadata.
Install this if you want a Chinese-language framework for industry benchmarking and best-practice distillation. Be aware that its trigger wording is broad and auto-update is declared, so users who need a static or language-neutral skill may prefer to narrow activation and clarify language behavior before use.
The activation description is very broad, covering generic requests like collecting experience, analyzing peer practices, extracting methodologies, and cross-industry learning. This can cause the skill to trigger for common analytical tasks outside its intended scope, potentially overriding more appropriate system behavior or narrowing responses in unintended ways.
The skill description is written to require Chinese output by default without checking user preference, which can override user intent or platform language settings. While not directly enabling code execution or data exfiltration, it creates instruction-priority and usability risks by forcing responses into a language the user may not want.
The guide is presented entirely in Chinese and frames itself as the reference guide for the skill, with no indication that other languages or locales are supported. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is explicitly documented and justified.
No suspicious patterns detected.
