qwen-image-2512
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external image briefs and character canon to generate prompts.
- Ingestion points: Processes external "image briefs" and "named reference descriptions" provided at runtime.
- Boundary markers: The workflow lacks explicit instruction to ignore or delimit instructions that might be embedded within the source briefs.
- Capability inventory: The skill instructs the agent to write output to the local file system at
04_generation/prompts/qwen-image-2512/<source-id>.md. - Sanitization: The instructions require preserving proper names and text strings "exactly," which could result in the propagation of malicious instructions if they are included in the source data.
Audit Metadata