decision-logic-framework

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFE
Full Analysis
  • PROMPT_INJECTION (SAFE): No malicious override patterns or safety bypass instructions were detected. The skill's instructions are focused on architectural organization.
  • DATA_EXFILTRATION (SAFE): No hardcoded credentials, sensitive file paths, or exfiltration patterns are present.
  • REMOTE_CODE_EXECUTION (SAFE): No remote scripts or unverified dependencies are executed. Documentation references common packages (httpx, click) as best-practice examples.
  • COMMAND_EXECUTION (SAFE): The framework defines criteria for when system commands should be used in scripts but does not execute any commands itself.
  • OBFUSCATION (SAFE): All content is in plain text (Markdown and TOON formats) with no hidden characters, encoded strings, or deceptive formatting.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 02:25 PM