brand-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No instructions designed to override agent behavior, extract system prompts, or bypass safety filters were detected.
- [Data Exposure & Exfiltration] (SAFE): The skill does not request access to sensitive files or credentials, and no network exfiltration patterns are present.
- [Remote Code Execution] (SAFE): The document contains no commands for downloading or executing remote code or scripts.
- [Obfuscation] (SAFE): The text is fully transparent with no use of Base64, zero-width characters, or other encoding methods.
- [Command Execution] (SAFE): While the text mentions python-pptx, it does not include any executable commands, subprocess spawning, or shell script logic.
Audit Metadata