The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill uses user-provided concepts as the primary source for generating HTML content. The instructions do not explicitly mandate sanitization or the use of boundary markers when the agent interpolates this user content into the generated HTML and JavaScript code.
Ingestion points: User input for concept, audience, and depth defined in SKILL.md.
Boundary markers: Not specified in the generation instructions.
Capability inventory: Writing and generating self-contained .html files containing embedded JavaScript.
Sanitization: No explicit requirement for escaping or validating input before interpolation.
[DYNAMIC_EXECUTION]: The skill generates functional code in the form of HTML and Vanilla JavaScript to support interactive elements like sliders and toggles.

explainer-infographic