The Agent Skills Directory

[SAFE]: The skill instructions and metadata do not contain any malicious patterns, such as prompt injections, obfuscation, or persistence mechanisms. The primary focus is on creative art generation and visual design.
[INDIRECT_PROMPT_INJECTION]: The skill incorporates user-provided creative inputs to define the 'soul' of the artwork. This represents a potential surface for indirect prompt injection; however, the skill prioritizes creative interpretation over literal instruction, and its capabilities are restricted to generating non-executable visual and text files, posing no significant security risk.
Ingestion points: User-provided subtle input or instructions described in SKILL.md.
Boundary markers: None explicitly defined in the prompt interpolation.
Capability inventory: Ability to generate .md, .pdf, and .png files; access to the local ./canvas-fonts directory.
Sanitization: No specific input sanitization steps are mentioned.
[SAFE]: The skill references a local canvas-fonts directory containing license files for various open-source fonts. This is a standard and safe practice for managing design assets and ensuring compliance with font licenses.

canvas-design