nano-banana-pro
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides templates and examples for interacting with the Refly platform using official CLI tools (
refly). These commands are used according to their documented purposes for workflow execution and file management. - [INDIRECT_PROMPT_INJECTION]: The skill illustrates a workflow that processes data from tool outputs (such as file IDs and names). While this constitutes a potential injection surface if an external actor could influence file metadata, it is a standard design pattern for the platform's automation capabilities and involves no immediate or obvious security risk.
Audit Metadata