The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to populate diagram templates, which creates a potential surface for injection if malicious content is supplied by the user. 1. Ingestion points: User requests and descriptions are used to fill {{TITLE}}, {{DESCRIPTION}}, and {{MERMAID_CODE}} placeholders in the generated MicroSim package files. 2. Boundary markers: The templates do not employ specific delimiters or warnings to separate user-provided content from the surrounding HTML and Mermaid logic. 3. Capability inventory: The skill has the capability to write directories and files to the local file system at /docs/sims/. 4. Sanitization: The instructions do not specify any validation, escaping, or sanitization of the user-provided data before it is interpolated into the templates.
[EXTERNAL_DOWNLOADS]: The generated main.html template includes a reference to the Mermaid.js library hosted on the well-known jsdelivr.net CDN. This is a legitimate and common practice for the skill's primary function of rendering interactive diagrams.

mermaid-generator