behavioral-modes

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE] (SAFE): No malicious behavior, data exfiltration, or obfuscation patterns were detected across the provided files.
  • [NO_CODE] (SAFE): The skill contains only documentation and instructional templates (markdown); no scripts or binaries are present.
  • [PROMPT_INJECTION] (SAFE): The behavioral modes (e.g., BRAINSTORM, IMPLEMENT) are benign task-specific instructions designed to optimize AI performance rather than bypass safety guardrails.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:38 PM