Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent markdown-only memory-management skill, with ordinary local persistence risks that users should understand before saving memories to disk.
Install only if you want a Chinese-language agent memory helper. Treat saved memory files as potentially sensitive, choose explicit storage paths, avoid saving secrets or private conversation content unless intended, and consider narrowing invocation to the exact skill name if your platform supports trigger customization.
The trigger keywords include extremely generic terms such as "memory," "smart," "manager," and "agents," which are likely to match ordinary user language and cause the skill to activate unintentionally. In a skill with read/exec capabilities and disk persistence features, accidental invocation increases the chance of unintended file operations or memory handling without clear user intent.
The skill advertises save/load functionality and configurable file paths, but it does not clearly warn users that conversation or memory contents may be written to disk and later reloaded. This creates a privacy and data-handling risk because users may unknowingly persist sensitive information locally, especially in agent environments handling personal or proprietary context.
The natural-language description is presented in Chinese while the surrounding manifest fields and platform context are otherwise not documented as China- or Chinese-specific. This creates a locale/language constraint without explicit user opt-in or justification, which can violate language-choice policy expectations.
No suspicious patterns detected.
