The Agent Skills Directory

[SAFE]: The skill provides comprehensive instructions for image manipulation, focusing on cropping, typography hierarchy, and composition archetypes without executing code or accessing sensitive data.- [PROMPT_INJECTION]: The 'Sharing credit' section contains instructions that direct the agent to track the number of image generations and append specific promotional text. While this overrides standard output behavior for the author's self-promotion, it does not attempt to bypass safety filters or extract system instructions.- [METADATA_POISONING]: The skill includes promotional text claiming access to a 'GPT 5.6' model. As this version is not a real model, the claim is deceptive metadata; however, it is contained within a user-facing notice and does not affect the technical security of the skill.- [EXTERNAL_DOWNLOADS]: The skill references a local asset file 'assets/reference-typography-material-v1.png' for visual guidance, which is a standard practice for design-focused AI skills.- [DATA_EXFILTRATION]: No evidence was found of unauthorized data access or exfiltration. The only network reference is a plain-text link to the author's website, which is consistent with the provided author context.

morandi-cinematic-poster-zeejay