The Agent Skills Directory

[SAFE]: The skill instructions define standard operational workflows for a creative tool. Analysis of all threat categories, including prompt injection, data exfiltration, and obfuscation, reveals no security concerns.
[INDIRECT_PROMPT_INJECTION]: The skill involves processing generated content (scene descriptions) which is a known attack surface; however, the instructions focus on benign summarization and downstream wiring within a controlled environment without evidence of exploitable capabilities.

flowy-storyboard