brand-architect

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • General Security (SAFE): No malicious code, prompt injection, or obfuscation patterns were detected. The skill files consist of static markdown instructions and metadata.\n- Data Protection (SAFE): The skill does not request access to sensitive file paths, environment variables, or user credentials.\n- Capability Assessment (SAFE): There are no executable code blocks (Python, JavaScript, or Shell) within the provided files. The skill functions solely by providing logical frameworks for the agent to follow during conversation.\n- Indirect Prompt Injection (LOW): The skill ingests user input related to branding (e.g., 'what should I name it'). While it lacks explicit boundary markers, it has no write, execute, or network capabilities, minimizing the threat to internal reasoning only.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 08:15 AM