The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for processing untrusted data by ingesting user-provided captions and assets into rendering tools.
Ingestion points: The skill instructions specify using "user's gameplay captures," "user's real gameplay captures," and "captions from render_text" derived from user input in SKILL.md.
Boundary markers: There are no specific boundary markers or escaping instructions provided for the text interpolation steps.
Capability inventory: The skill utilizes render_text, compose_image, text_to_image, and render_video_timeline to process these inputs.
Sanitization: No explicit sanitization or validation of the user-provided text content is mentioned before it is passed to rendering tools.
[COMMAND_EXECUTION]: The skill documentation includes an instruction for the user to install sibling components using a CLI tool.
Evidence: The text recommends the command npx skills add layerai/skills --skill <name> to the user if dependencies are missing.
Analysis: This reference is consistent with the skill's own vendor (layerai) and represents a manual installation step for the user rather than an automated or hidden execution pattern.

layer-store-assets