minimax
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- COMMAND_EXECUTION (SAFE): The skill uses curl to make HTTP requests to the MiniMax API. While it uses
bash -cto ensure environment variables are handled correctly, the commands are transparent and target legitimate service endpoints.\n- DATA_EXFILTRATION (SAFE): Network activity is directed only to the official MiniMax API domains (api.minimax.io and api.minimaxi.chat) as required for the skill's primary function.\n- CREDENTIALS_UNSAFE (SAFE): The skill correctly uses thevm0_secretsmetadata to manage theMINIMAX_API_KEY, ensuring that sensitive credentials are not hardcoded or exposed in the skill content itself.
Audit Metadata