scenario-minimax-video
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user inputs including text prompts and reference media (images, videos, and audio) which are then used as parameters for the
model_runtool. This creates a surface for indirect prompt injection where instructions embedded in media metadata or the prompt itself could influence agent behavior. - Ingestion points:
prompt,referenceImages,referenceVideos, andreferenceAudiofields inSKILL.md. - Boundary markers: No specific delimitation or 'ignore instructions' markers are defined for these inputs.
- Capability inventory: The skill uses
model_run,upload_asset, andsearchtools. - Sanitization: No explicit sanitization or validation of the input content is described.
- [EXTERNAL_DOWNLOADS]: The documentation suggests installing additional skills using
npx skills add scenario-labs/skills. This command refers to the official repository of the skill author (scenario-labs) and is considered a safe vendor resource reference.
Audit Metadata