grad-brand-equity

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional and provides a framework for brand analysis. It does not perform any operations outside of processing provided text data.
  • [PROMPT_INJECTION]: No patterns associated with prompt injection, such as instruction overrides, safety bypasses, or system prompt extraction, were detected in the instructions or examples.
  • [DATA_EXFILTRATION]: No network operations (curl, wget, fetch) or access to sensitive file paths (e.g., credentials, ssh keys) were identified. The data analyzed in the examples is fictional marketing data.
  • [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or dynamic execution patterns. It relies entirely on standard LLM reasoning.
  • [EXTERNAL_DOWNLOADS]: There are no remote scripts, package installations, or external dependencies requested or utilized in the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 07:10 AM