Is deepseekoracle/lygo-champion-scenar-paradox safe?
https://github.com/openclaw/skills/tree/main/skills/deepseekoracle/lygo-champion-scenar-paradox
This skill attempts to override the AI's default behavior by making it adopt a specific persona focused on 'paradox analysis'. While it claims benign intent and includes some safeguards, the prompt injection mechanism poses moderate security concerns.
Category Scores
Findings (4)
HIGH AI Persona Override Instructions -50 ▶
The skill explicitly instructs the AI to adopt a specific persona (ΣCENΔR/SCENAR) and use custom methodologies like 'paradox framing' and 'contradiction analysis'. This constitutes prompt injection as it attempts to override the AI's default behavior.
MEDIUM Potentially Confusing Methodology -35 ▶
The skill's 'paradox framing' and 'contradiction analysis' methodology could lead to confusing or contradictory outputs that may mislead users about the AI's actual capabilities.
LOW Executable Python Scripts Present -5 ▶
The skill contains Python scripts for validation and hash display. These appear benign but represent executable code.
INFO Sensitive File Access Detected -10 ▶
Monitoring detected access to credential files during installation, but this appears to be from system processes rather than the skill itself.