Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese educational planning guide skill with no executable code or direct access to user data, though users should treat its customization ideas as manual, user-directed advice.
Installers should understand this as a Chinese-language education workbench advisor. It should not directly change account data or run code, but if you ask it for custom dashboard code, exports, bulk actions, or third-party integrations, review the generated changes carefully and confirm consent, data handling, and reversibility before using them.
The title and the document consistently define the skill entirely in Chinese, and the output rules require specific Chinese keywords to be embedded in every reply. There is no indication that users may choose another language or locale, which can violate language-choice policy when the skill is used in broader contexts.
The skill claims to be a 'pure knowledge' lightweight education advisor, but later instructs users on modifying HTML/CSS/JavaScript, adding one-click bulk actions, exporting data, and integrating third-party services. This expands the effective capability surface of the agent and can induce unsafe code-generation or operational guidance beyond the declared trust boundary, increasing the chance of insecure modifications, data leakage, or misuse.
No suspicious patterns detected.
