The Agent Skills Directory

[SAFE]: The skill provides instructions for using standard platform tools like generate_video, get_timeline, and get_media to fulfill user requests for video content. No malicious patterns, obfuscation, or unauthorized access attempts were detected.
[SAFE]: The skill demonstrates security awareness by including explicit guardrails against generating likenesses of real individuals without consent and maintaining brand-safe content.
[SAFE]: The skill processes user-supplied descriptions and dialogue (ingestion points) for interpolation into video generation tools. It utilizes structured motion-prompt formulas (boundary markers) to control these inputs. The platform capabilities involved include video/audio generation and timeline manipulation (capability inventory). While explicit sanitization logic is not provided in the instructions (sanitization), the use of specific prompt templates and guidance to inform users of costs (e.g., for Seedance 2) follows safe operational practices.

ugc-video-prompts