brand-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- NO_CODE (SAFE): The skill consists solely of markdown text and YAML metadata. No scripts, binaries, or automated tasks are included.
- Unverifiable Dependencies & Remote Code Execution (SAFE): While the text mentions the 'python-pptx' library, the skill does not attempt to install, download, or execute any external packages or remote code.
- Prompt Injection (SAFE): The instructions are strictly limited to styling guidelines and do not contain any patterns aimed at overriding agent behavior or bypassing safety filters.
- Data Exposure & Exfiltration (SAFE): No file paths, credentials, or network operations are present in the provided file.
Audit Metadata