brand-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- Prompt Injection (SAFE): The content consists of stylistic guidelines and does not contain any commands designed to override AI safety protocols or system instructions.
- Data Exposure & Exfiltration (SAFE): No credentials, private keys, or sensitive file paths are present. There are no network requests or data transfer commands.
- Obfuscation (SAFE): The file contains clear, readable markdown text with no hidden characters, Base64 encoding, or homoglyphs.
- Unverifiable Dependencies (SAFE): The skill mentions 'python-pptx' in its technical description but does not contain a requirements file or installation commands. It appears to be documentation for a capability rather than a script that installs dependencies.
- Indirect Prompt Injection (LOW): The skill is intended to process 'artifacts' to apply styling. While this creates a data ingestion surface, the skill itself provides no mechanism for command execution, and the instructions are limited to visual formatting.
Audit Metadata