video-generation
Pass
Audited by Gen Agent Trust Hub on Apr 30, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill bundles several Python scripts (e.g.,
text_to_video.py,image_to_video.py) that are executed via the command line. These scripts useargparseto safely handle user-provided inputs such as prompts and file paths. - [EXTERNAL_DOWNLOADS]: The scripts interact with Google's GenAI API to generate video content. Upon completion, the scripts download the generated
.mp4files from Google's infrastructure to the user's local directory. - [CREDENTIALS_UNSAFE]: The skill implements a secure lookup pattern for the
GEMINI_API_KEY, checking environment variables and local.envfiles. It does not contain hardcoded secrets and provides clear instructions for users to manage their keys safely. - [REMOTE_CODE_EXECUTION]: All network operations are directed towards official Google GenAI endpoints using the
google-genailibrary. No patterns of arbitrary code execution or untrusted remote script fetching were detected.
Audit Metadata