Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward code-review skill with no executable payload, persistence, hidden data flow, or privilege escalation instructions.
Before installing, expect this skill to guide code reviews and produce prioritized Markdown findings, likely in Chinese unless the user asks otherwise. For sensitive repositories, explicitly state the files or modules to review.
The trigger description is very broad (e.g., any request about code review, code quality, code checking, optimization) without clear boundaries on repository scope, file access, or task limits. In an agent environment, this can cause over-invocation or application to unintended inputs, increasing the chance of excessive data exposure or unsafe autonomous actions on sensitive codebases.
The skill content is entirely specified in Chinese and describes the deliverable and workflow only in that language, without indicating that users may choose another language. This creates a locale/language constraint without explicit opt-in or documented justification.
No suspicious patterns detected.
