SwitchHat: Multi-Persona Prompt Engine

LLMs are trapped in a predictable, single-mode personality that stifles critical reasoning and creative depth. SwitchHat injects specialized cognitive frameworks into prompts to force models out of their default bias.