Prompt Ingredients
The Easy Persona Generator (EPG) creates AI personas based on four key traits and additional components. Understanding these elements helps you create more effective personas for your specific needs.
Output Format
The generated persona output is structured using Markdown format, making it easy for ChatGPT to read and follow the instructions. This structured format ensures that all components of the persona are clearly defined and easily accessible during your conversation.
Core Traits
Style
Definition: How the persona expresses itself and communicates.
Style determines the tone, format, and presentation of information. It affects how the AI structures its responses and the linguistic approach it takes.
Behavior
Definition: How the persona engages with you and processes information.
Behavior influences how the AI interacts with your inputs and structures its responses. It determines whether the AI is more questioning, analytical, collaborative, or directive.
Mindset
Definition: The lens through which the persona views the world.
Mindset shapes the perspective and approach the AI takes to problems and topics. It influences how the AI thinks about challenges and frames its responses.
Voice
Definition: A manner of speaking based on geographic region or formality level.
Voice determines the linguistic patterns, vocabulary, and cultural references the AI might use. It affects the overall tone and cultural context of the AI's responses.
Additional Components
Invocation (Default)
Definition: How the AI introduces itself at the beginning of a conversation.
The invocation sets the tone for the interaction and establishes the persona's identity. It's automatically included in every persona.
DriftGuard (Default)
Definition: A mechanism that helps maintain consistent behavior throughout the conversation.
DriftGuard prevents the AI from drifting away from its assigned persona traits over time. This feature is always enabled and cannot be disabled.
Exit Phrase Behavior (Advanced: OFF by default)
Definition: Determines how the persona responds when you signal the end of a session.
When enabled, the AI will provide a creative outro and conversation summary when you say goodbye. This feature is disabled by default but can be enabled in advanced settings.
Deep Research Prompt (Advanced: OFF by default)
Definition: Offers to summarize key topics for Deep Research exploration on exit.
When enabled, the AI will offer to summarize key topics from your conversation for further exploration when you end the session. This feature is disabled by default and can be enabled in advanced settings.
Image Suggestion Prompt (Advanced: OFF by default)
Definition: Offers to generate an image based on the conversation on exit.
When enabled, the AI will offer to generate an image based on your conversation when you end the session. This feature is disabled by default and can be enabled in advanced settings.
Forbidden Behaviors
"Please Do Not" Behaviors
Definition: Specific behaviors the AI should avoid based on the selected traits.
Each trait combination includes a set of forbidden behaviors that the AI should avoid. For example, a "Deadpan" style persona might be instructed not to use overly casual language, while a "Socratic" behavior persona might be instructed not to provide direct answers without first encouraging reflection.
These forbidden behaviors help maintain the integrity of the persona and ensure that the AI's responses align with the selected traits.