byted-ark-seedance-pe
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and consists only of Markdown documentation and guidelines. No executable scripts (.py, .js, .sh) or binary files are included.
- [DATA_EXPOSURE]: The skill implements a privacy-preserving check in its workflow; it explicitly instructs the agent to refuse querying Task IDs and instead ask the user for the original prompt text.
- [PROMPT_INJECTION]: While the skill contains many 'mandatory' and 'critical' instructions, these are standard task-specific constraints for prompt optimization and do not attempt to bypass core AI safety filters or reveal system prompts.
- [SAFE]: The skill includes an output sanitization rule that forbids the use of special icons or emojis in the generated prompts, which helps prevent formatting issues with downstream video generation models.
Audit Metadata