seedance-prompt-short
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or security risks were identified in the skill instructions. The skill focuses on prompt engineering and compression logic without requiring high-risk capabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied briefs and sequence data. While it handles untrusted data, it lacks the capability to perform dangerous operations like network requests, file system writes, or command execution, effectively neutralizing the risk of indirect injection exploitation. Evidence of data ingestion is found in the 'Compression Priority' and 'Compact Templates' sections where user briefs are parsed and reformatted.
Audit Metadata