seedance-epic-fantasy-scifi
Warn
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: MEDIUMOBFUSCATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The file
references/cases.mdcontains repeated sequences of the Unicode characterU+FE0F(Variation Selector-16) following the#AIarthashtag in cases E1, E3, and E7. These non-visible characters match patterns for data obfuscation through special Unicode sequences.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to take user-provided subjects and settings and combine them with external reference cases to create prompts, creating a vulnerability surface for indirect prompt injection.\n - Ingestion points: User-provided descriptions for the video subject and setting, as well as the scraped content in
references/cases.md.\n - Boundary markers: Absent. The skill instructions do not specify the use of delimiters or instructions for the agent to ignore potential commands within user data.\n
- Capability inventory: The skill utilizes the agent's text generation and drafting capabilities.\n
- Sanitization: Absent. The skill does not include any steps to validate or escape external content or user inputs before processing.
Audit Metadata