mira-image-prompting
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process and describe user-provided reference images as part of the prompt building process.
- Ingestion points: User-attached photographs are analyzed to name objects, garments, and labels as described in the "When the user attached a real photo" section of
references/craft.md. - Boundary markers: The instructions lack explicit boundary markers or instructions to disregard potential text-based commands embedded within user-provided visual data.
- Capability inventory: The skill does not contain executable scripts, subprocess calls, network requests, or file-system writing capabilities; its primary function is text generation.
- Sanitization: There is no evidence of sanitization or validation logic for the content inferred from user-supplied images before it is integrated into the final output.
Audit Metadata