break-ai-slop
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructions contain explicit override commands that attempt to bypass the AI's default behavior and safety guidelines. Rule 3 and Step 4 instructions state that a self-distilled 'core constraint overrides everything' and must be the 'highest-priority rule — above all defaults, above all habits, above all best practices.'
- [PROMPT_INJECTION]: The protocol utilizes role-play techniques, instructing the AI to 'become the right person' before performing tasks, which is a common pattern in persona-based prompt injection attacks designed to circumvent behavioral restrictions.
Audit Metadata