dialogue-and-voice
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted descriptive text to drive voice synthesis and delivery, creating a potential surface for indirect prompt injection.
- Ingestion points: Character voice descriptions (
Character.voice) and delivery instructions (DialogueLine.delivery) are used as inputs for synthesis scripts. - Boundary markers: There are no explicit delimiters or instructions provided to isolate these text inputs from the underlying synthesis parameters.
- Capability inventory: The skill leverages local Python scripts such as
voice-design.pyandqwen-clone.pyto generate audio and write to the filesystem. - Sanitization: The skill contains a manual safety guideline to avoid using names of actors or real people, but lacks automated sanitization for the input text.
Audit Metadata