Security audit
Security checks for vulnerabilities and agentic risk
This skill does what it says: it generates an editable draw.io link from diagram text, with only minor usability and scoping caveats.
Before installing, know that the skill may run a bundled Python script to create a diagrams.net link and that opening the link sends the diagram content to app.diagrams.net in your browser. Avoid putting sensitive information in diagrams unless you are comfortable opening it in that external service.
The skill instructs the agent to execute a Python script that generates a draw.io URL, which implies code execution and likely external network interaction, yet the skill declares no explicit tool scope or permissions. This increases the risk of unintended script execution or network access without clear sandboxing and makes the skill harder to review and constrain safely.
The trigger description is extremely broad, covering essentially any request for a visual representation, which can cause the skill to activate in many unrelated contexts. Over-broad routing increases the chance that untrusted user content is funneled into script execution paths or that this skill overrides more appropriate, narrower skills with safer handling.
The skill mandates outputting a markdown link with Chinese anchor text regardless of user preference or locale. While not a direct code-execution flaw, forced language output can mislead users, reduce transparency about the destination, and degrade trust and usability in security-sensitive contexts where clear user understanding matters.
The printed markdown link text is hard-coded as Chinese (点击查看图表), which imposes a specific language on all users. The file does not provide any language selection, opt-in, or documentation justifying a locale-specific constraint.
No suspicious patterns detected.
