Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed Chinese thesis-review template that reads thesis content, searches literature context, and generates review comments plus a Word document.
Before installing, consider whether your thesis content is confidential, because the workflow may search the web with topic keywords and create a Word review file in the working directory. If you only want chat feedback or another language, ask the agent explicitly before using the skill.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
---
name: thesis-review
version: 2.1.0
description: "Review academic degree theses with structured evaluation framework supporting bachelor/master/PhD..."
The skill instructs the agent to automatically generate and save a .docx file in the working directory as a mandatory step, but it provides no requirement to obtain explicit user confirmation, disclose the destination, or warn about filesystem side effects. In an agent environment, silent file creation can lead to unexpected local writes, overwriting of existing files, or abuse in chained workflows where the user only asked for analysis text output.
The trigger list is broad and maps to very common academic phrases such as '论文评审' and '毕业论文', which can cause the skill to activate for general thesis-help requests rather than only explicit review tasks. In an academic assistant context, this creates scope ambiguity and may unexpectedly route unrelated user prompts into this skill, increasing the chance of inappropriate behavior or misleading outputs.
This markdown file is written as a skill reference entirely in Chinese from the title onward, with no indication that users may choose another language or that the skill is intentionally limited to a Chinese-language context. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
The entire template is written in Chinese and provides 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.