gws-modelarmor-sanitize-prompt
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a clean interface for the 'gws' CLI tool to sanitize user-provided text through Google Cloud Model Armor templates. It does not include any executable scripts, remote downloads, or obfuscated instructions.
- [COMMAND_EXECUTION]: The skill uses the 'gws' local binary to perform its tasks. The command usage is structured correctly and does not involve risky patterns like shell piping to interpreters or privilege escalation.
- [DATA_EXPOSURE]: No hardcoded credentials, API keys, or access to sensitive local file paths (like .ssh or .aws) were found. The skill handles user prompts for the purpose of safety filtering, which is its intended primary purpose.
Audit Metadata