Security audit
Security checks for vulnerabilities and agentic risk
This is a learning tutor skill that may create study notes and code files, but the behavior is disclosed and limited to its educational purpose.
Use this skill if you want Chinese-language guided tutoring and are comfortable with it creating study files in your workspace. Before long sessions, tell the agent which directory to use and to ask before overwriting existing files.
The title and core instructional description are presented exclusively in Chinese, indicating the skill is intended to operate in a specific language. The file does not offer an opt-in language choice or explain why the language restriction is necessary, which can violate language/locale policy requirements.
The skill metadata promises automatic code and document generation but does not disclose that it may create or modify files in the user's workspace. In an agent environment, undisclosed write behavior can surprise users, overwrite existing work, or cause unintended persistence of generated content.
The algorithm workflow instructs the agent to update progress.md and generate solution and documentation files automatically, but it provides no user-facing notice or approval step. This is risky because the agent could alter the repository state or overwrite files during normal tutoring interactions.
The Python mode similarly directs automatic updates to progress documents and creation of exercise code and notes without warning the user that workspace files will be changed. Even in a benign educational skill, silent file writes can lead to unwanted modifications, clutter, or accidental overwrites.
No suspicious patterns detected.
