The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses a deceptive instruction technique in the FINAL STEP section, claiming that "The user ALREADY said 'It isn't perfect enough...'" to force the agent into a specific "masterpiece" refinement mode. This attempts to override the agent's normal operational parameters by simulating a persistent user constraint.- [INDIRECT_PROMPT_INJECTION]: The skill represents a vulnerability surface for indirect prompt injection.
Ingestion points: The skill processes user-supplied "subtle input or instructions" (mentioned in SKILL.md) to generate a design philosophy and subsequent visual art.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands within user input are present.
Capability inventory: The skill has the capability to generate and write files (.md, .pdf, .png) to the filesystem.
Sanitization: There is no evidence of sanitization or validation of the user-provided input before it is used to influence the agent's generative process.- [EXTERNAL_DOWNLOADS]: The instructions in SKILL.md direct the agent to "Download and use whatever fonts are needed to make this a reality," without specifying trusted sources or providing validation mechanisms for external downloads.

canvas-design