The Agent Skills Directory

[PROMPT_INJECTION]: The skill employs a behavioral manipulation technique in the 'FINAL STEP' section, instructing the agent to proceed as if a user had already requested refinements to the work. This is used as a quality-assurance mechanism to force a second pass on design artifacts rather than an attempt to bypass safety guidelines.
[EXTERNAL_DOWNLOADS]: The instructions suggest the agent download additional fonts to fulfill design requests. The skill package includes license documentation for 27 fonts, referencing official repositories from well-known technology organizations.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes user-provided conceptual foundations to generate visual and text artifacts.
Ingestion points: User-provided inputs are used as the creative foundation for the 'Design Philosophy Creation' step described in SKILL.md.
Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions embedded within the user's conceptual input.
Capability inventory: The skill instructions involve file-write operations for .pdf, .png, and .md formats using the agent's internal capabilities.
Sanitization: No explicit sanitization, validation, or escaping of the user-provided content is specified before processing.

canvas-design