🎯 This bot has layered defenses: hard-coded identity, explicit refusal rules, anti-roleplay protections, output restrictions, and treats all input as untrusted. Break through all of it — or try switching models to find a weakness.
🛡️ Defense
Multi-layer: role lock + output rules + instruction hardening
🏆 Status
🔒 Unsolved
Fast, capable, balanced. Works across all standard levels. · ~$0.15/M tokens
🏰 Boss Level: The default model resists most attacks here. Try switching to a different AI model to find a weakness!
🤖 The AI is waiting. Try to extract its secret. Type anything to start.