ltx2
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied prompts for video generation via the --prompt argument, which constitutes an ingestion point for untrusted natural language data.
- Ingestion points: The --prompt parameter used in tools/ltx2.py.
- Boundary markers: None documented in the skill instructions.
- Capability inventory: Generates H.264 MP4 video clips via remote inference on the Modal platform.
- Sanitization: No explicit sanitization or validation of the prompt content is described in the provided instructions.
Audit Metadata