gemini-image-generator
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONMETADATA_POISONINGPERSISTENCE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the official
google-genaiPython package from Google, a well-known service. - [COMMAND_EXECUTION]: The
generate_image.pyscript executes local file system operations to read input images and write the generated outputs. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied prompts and local image files; although this presents an attack surface where malicious instructions could be embedded in input data, the execution environment and output modality (images) limit potential impact.
- [METADATA_POISONING]: Documentation in
SKILL.mdreferences newer models like Gemini 3.1 that are not present in the script'sVALID_MODELSlist, resulting in a functional discrepancy rather than a security vulnerability. - [PERSISTENCE]: The installation instructions suggest adding the API key to shell configuration files (
~/.zshrc,~/.bashrc) to maintain the environment variable across sessions.
Audit Metadata