The Agent Skills Directory

[SAFE]: The skill provides structured guidelines and prompt templates for image generation. It lacks any executable scripts, obfuscated content, or instructions designed to circumvent AI safety guardrails.- [DATA_EXFILTRATION]: The skill references reading local files such as strategy/brand-knowledge-base.json and strategy/product-truth/{product}.md to populate context. This behavior is standard for retrieval-augmented generation within a local environment and does not pose an exfiltration risk.- [REMOTE_CODE_EXECUTION]: The skill mentions a Python script enrich_image_prompt.py but does not facilitate its download from external sources or its execution with untrusted parameters.

image-style-registry