vidu-video
Pass
Audited by Gen Agent Trust Hub on Jul 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust security model for API interaction, requiring explicit operator approval and validation of request digests before any network communication or expenditure occurs.
- [SAFE]: The provided Python implementation follows secure coding practices, utilizing only standard libraries and employing atomic file writes with restricted permissions (0o600) to manage local execution state.
- [EXTERNAL_DOWNLOADS]: The skill is designed to communicate with the Vidu API (api.vidu.com) to generate video content. This is consistent with its primary purpose and targets a well-known service.
- [DATA_EXFILTRATION]: While the skill transmits data to an external provider, it correctly avoids hardcoding credentials, instead relying on environment variables. The instructions explicitly warn against placing sensitive tokens in logs or analytics.
- [COMMAND_EXECUTION]: The script uses standard Python library functions to perform network requests and manage local files. These operations are scoped strictly to the functional requirements of the API integration.
Audit Metadata