nano-banana-2
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
infshCLI andinferenceshPython SDK to execute image generation tasks. These tools are part of the vendor's own infrastructure. - [EXTERNAL_DOWNLOADS]: References an installation script for the CLI tool from the vendor's GitHub repository and suggests adding related skills via
npx. - [PROMPT_INJECTION]: The skill processes external data which could lead to indirect prompt injection.
- Ingestion points: User-provided
promptstrings and externalimagesURLs processed by the image generation model (SKILL.md). - Boundary markers: None identified in the provided examples to separate instructions from untrusted data.
- Capability inventory: Uses the
infshtool to perform remote model inference. - Sanitization: No input sanitization or prompt wrapping is evident in the instructions.
Audit Metadata