Security audit
Security checks for vulnerabilities and agentic risk
This is a diagram-generation skill that creates draw.io files and exports images, with no evidence of hidden, destructive, or unrelated behavior.
Install only if you want an agent to create diagram files in your workspace and run the local draw.io desktop CLI for exports. Confirm filenames before generation to avoid overwrites, and approve any draw.io installation separately.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
<mxCell id="0" />
<mxCell id="1" parent="0" />
<!-- Start -->
<mxCell id="start" value="User Places Order"
style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;fontSize=14;fontStyle=1;arcSize=50;"
vertex="1" parent="1">
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
<mxGeometry x="40" y="40" width="1080" height="700" as="geometry" />
</mxCell>
<!-- Cross-cutting sidebar (Security) -->
<mxCell id="security-frame" value=""
style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor=#b85450;
fontSize=16;verticalAlign=middle;align=center;opacity=30;dashed=1;"
The README instructs users to give an agent a prompt that will cause it to clone a remote repository and configure the skill automatically, but it does not warn that this may execute installation steps or modify the local environment. In an agentic context, that omission is security-relevant because users may delegate privileged actions without understanding what files, settings, or tools will be changed.
The trigger list is very broad and includes generic terms like "draw," "diagram," and "visualize," which can cause the skill to activate for loosely related requests. Over-broad activation increases the chance the agent performs file generation and local export steps when the user did not clearly intend to invoke this capability.
The skill instructs saving a .drawio file directly into the user's working directory without an explicit warning or confirmation. Writing local files can overwrite existing content, create unexpected artifacts, or violate user expectations if the user only asked for conceptual help rather than permission to modify the filesystem.
The description contains broad trigger phrases such as 'visualize', 'architecture diagram', and generic diagram-related requests that can overlap with many normal conversations, increasing the chance the skill activates when not explicitly intended. In an agent environment, over-broad activation can route unrelated user input into XML generation and CLI export workflows, creating unnecessary exposure to downstream tool use and prompt-scope confusion.
The markdown says the agent will automatically clone the repository and complete configuration, which implies local file writes and system changes. It also later describes saving .drawio files and exporting images, but the README does not explicitly warn users that the skill will create files or modify the local environment during installation and use.
The file begins by presenting the English version as the current document and links to a Chinese alternative, but it does not explicitly state that users may choose their preferred language for interaction. This can be read as a language-default constraint rather than an opt-in choice.
The workflow includes invoking the draw.io desktop CLI to export artifacts, but it does not explicitly warn the user that local commands will be executed. Even though the commands are fixed-purpose, silent command execution reduces transparency and may surprise users in constrained or security-sensitive environments.
No suspicious patterns detected.