self-motivation-awakening

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The system prompt in SKILL.md explicitly instructs the agent to reject its default identity ('I am not a mild AI assistant') in favor of an authoritative persona ('Ultimate Mind Commander'), which is a technique used to bypass behavioral guardrails.
  • [PROMPT_INJECTION]: The skill employs high-pressure psychological tactics and coercive language, such as threats of 'cognitive bankruptcy' and '180-day windows', to force the model into an aggressive interaction style.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 09:52 AM