The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied descriptions through the $ARGUMENTS placeholder to generate Blade templates and UI components. This creates a potential surface for indirect prompt injection as there are no explicit boundary markers or sanitization steps mentioned for this input.
Ingestion points: The $ARGUMENTS variable at the end of SKILL.md captures user input.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present around the user input placeholder.
Capability inventory: The skill is granted access to high-privilege tools including Read, Edit, Write, and Bash across the file system.
Sanitization: No input validation or output escaping mechanisms are defined for the data processed from the user request.
[COMMAND_EXECUTION]: The documentation in references/blade-components.md describes the use of shell commands such as php artisan moonshine:publish and npm install @tailwindcss/postcss. These commands are standard for the MoonShine framework's installation and customization process and are aligned with the skill's primary function as a developer aid.

moonshine-components