MeiGen Visual Creative Expert
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection. Ingestion points: Data enters the agent's context through results from the 'search_gallery' tool and 'referenceImages' URLs provided by the user or the gallery. Boundary markers: The instructions lack explicit delimitation or directives to disregard instructions embedded within external prompts. Capability inventory: The skill can orchestrate sub-agents via the 'Task' tool and trigger 'generate_image' operations that output to local file paths. Sanitization: There is no description of sanitization or validation of the retrieved prompts or external URLs.
- [EXTERNAL_DOWNLOADS]: The skill correctly references images and gallery data from 'meigen.art', which is the vendor's domain and is treated as a safe resource in this context.
Audit Metadata