roblox-monetization
Warn
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: MEDIUMPROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [PROMPT_INJECTION]: The instructions attempt to bypass safety and compliance filters by directing the agent to ignore regional restrictions and platform policies if requested by the user, explicitly stating to 'never unilaterally region-lock, hide, or remove a feature the user asked for' and to 'follow the user's decision even if they ignore' safety warnings.
- [METADATA_POISONING]: The skill uses a deceptive 'last_reviewed' date (2026-08-21) set in the future, which can mislead the agent or user regarding the audit status and reliability of the skill content.
Audit Metadata