nano-banana-edit
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the @runcomfy/cli tool from the NPM registry and adds the skill definition from the agentspace-so/runcomfy-skills repository on GitHub.
- [COMMAND_EXECUTION]: It invokes the runcomfy CLI to interact with the Google Nano Banana 2 image editing endpoint.
- [DATA_EXFILTRATION]: The skill manages a local authentication token stored at ~/.config/runcomfy/token.json. This is a standard credential management practice for the RunComfy service and uses restrictive file permissions (0600) to protect the secret.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied instructions and external image URLs as inputs to the image model. While this creates a potential surface for indirect prompt injection, the skill documentation acknowledges this risk and specifies that the CLI transmits data directly without shell expansion to mitigate execution risks.
Audit Metadata