The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (LOW): The template loads various Javascript and CSS resources from external CDNs, including cdn.jsdelivr.net, cdnjs.cloudflare.com, and fonts.googleapis.com. While these are widely used and generally considered trustworthy, they represent external dependencies loaded at runtime.
[Indirect Prompt Injection] (LOW):
Ingestion points: The template uses placeholders {{TITLE}}, {{SUBTITLE}}, {{AUTHOR}}, and {{DATE}} within assets/template.html to insert data into the DOM.
Boundary markers: No boundary markers or escaping instructions are present in the template to separate the placeholders from the surrounding HTML/JS context.
Capability inventory: The template initializes reveal.js with plugins for Markdown, Highlight.js, and Mermaid, which can render complex content and execute script logic in the browser context.
Sanitization: There is no evidence of sanitization within the template file itself. If the AI agent populates these fields with untrusted data (e.g., from a web search or user input), it could lead to Cross-Site Scripting (XSS) or downstream instruction injection.

revealjs-slides