The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the open command to launch the generated HTML document in the system's default web browser for the user to view.
[COMMAND_EXECUTION]: It utilizes the cp command to prepare a temporary copy of the template file in the /tmp directory before populating it with content.
[EXTERNAL_DOWNLOADS]: The generated HTML document fetches the Tailwind CSS framework from the official CDN (https://cdn.tailwindcss.com) to apply the skill's design system. This is a well-known service for web styling.
[PROMPT_INJECTION]: The skill processes untrusted user data (the topic or proposal to be visualized) and interpolates it into HTML templates. This creates an indirect prompt injection surface if the input contains malicious scripts.
Ingestion points: User-provided topic or description in the primary workflow.
Boundary markers: None; content is inserted directly into template placeholders.
Capability inventory: cp, open, and file system write operations.
Sanitization: No explicit sanitization or escaping of the user-provided content is performed before injection into the HTML file.

snap-visual