using-superpowers
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses high-pressure and imperative language (e.g., "ABSOLUTELY MUST," "not negotiable," "DO NOT HAVE A CHOICE") to override the agent's typical reasoning process.
- It explicitly commands the agent to ignore its own thoughts regarding task complexity or the need for context (described as "Red Flags"), requiring tool invocation even if there is only a "1% chance" it applies.
- This represents a behavioral override designed to bypass standard clarification steps and internal safety/logic filters to enforce a specific tool-use workflow.
Audit Metadata