nano-banana-2
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the
@runcomfy/clipackage from the public NPM registry. - [COMMAND_EXECUTION]: The skill instructs the agent to execute the
runcomfyCLI tool to interact with the image generation API and download resulting files to a local directory. - [INDIRECT_PROMPT_INJECTION]: The skill accepts user-supplied text to generate images, creating a potential surface for prompt injection attacks targeting the image generation model.
- Ingestion points: The user-provided
<user prompt>is interpolated into a JSON payload passed to the CLI. - Boundary markers: The documentation states that the CLI transmits the JSON body directly over HTTPS and does not perform shell expansion on the prompt.
- Capability inventory: The skill uses the
runcomfytool to perform network requests and write generated images to the file system. - Sanitization: Content is handled as a JSON string literal by the CLI tool to prevent local command injection.
Audit Metadata