> ## Documentation Index
> Fetch the complete documentation index at: https://help.thedouble.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Persona

Overview

Persona lets you customize how your AI Double thinks, speaks, and responds — so every conversation feels authentically you.

Go to your **Double → Persona** to get started.

**What you can customize**

* Custom instructions
* Greeting message
* Tone & Personality (Mindset + Communication style)
* Topic flexibility
* Auto suggested replies
* Response structure
* Behavior when Double doesn't know something

**Custom Instructions**

Set permanent rules for how your Double behaves across every conversation.

Type your instructions in the text box — e.g. "Be very formal", "Use simple analogies", "Keep answers concise but friendly." You have up to 1500 characters.

**Tip:** These instructions are added to every prompt. Use this to set permanent rules for your AI's behavior.

**How to write good custom instructions**

Be specific and direct. A few examples:

* "Always respond in bullet points when listing more than 3 items"
* "Never discuss competitors by name"
* "Always end with a question to keep the conversation going"

Custom instructions take priority over tone presets — if there's a conflict, instructions win.

### **Greeting**

Set the very first message a visitor sees when they open your chat.

Type your greeting in the **Static greeting** field — e.g. "Hello! How can I help you today?"

**Tip:** Keep it short and inviting. A good greeting sets the tone and signals to visitors what they can ask.

### **Tone & Personality**

Choose how your Double comes across in conversation. Pick one of five presets:

| **Preset**       | **Description**                                                                                        |
| ---------------- | ------------------------------------------------------------------------------------------------------ |
| **Default**      | Balanced and versatile — works well for most use cases with a professional <br />yet approachable tone |
| **Friendly**     | Warm and conversational — like chatting with a helpful friend who genuinely <br />wants to connect     |
| **Warm**         | Supportive and empathetic — focused on making people feel heard, <br />valued, and understood          |
| **Professional** | Polished and authoritative — clear, confident communication focused on <br />delivering value          |
| **Quirky**       | Creative and playful — high-energy personality with humor, wit, and a touch<br />of the unexpected     |

Each preset sets the starting values for **Mindset** and **Communication style**. Expand either section to fine-tune individual variables.

To fine-tune further, edit individual variables — the preset will show **Custom configuration active**. Click **Reset variables to default preset** to go back to the original settings.

### **Mindset**

Controls how your AI Double thinks and approaches conversations. Expand the **Mindset** section to adjust two variables:

### **Confidence**

**Slider range:** Careful → Bold

Controls how assertively your Double makes statements and recommendations.

* Careful end: more measured, hedged responses
* Bold end: direct, confident language (e.g. "I'd recommend this approach — it's a solid choice.")

### **Curiosity / Questioning**

**Slider range:** Reserved → Inquisitive

Controls how much your Double probes and explores topics with visitors.

* Reserved end: answers without asking follow-up questions
* Inquisitive end: actively digs deeper (e.g. "Here's a breakdown — would you like to explore any part further?")

### **Communication Style**

Controls how your AI Double speaks and expresses itself. Expand the **Communication style** section to adjust four variables:

### **Tone**

**Slider range:** Casual → Corporate

Sets the register of your Double's language.

* Casual end: relaxed, conversational (e.g. "I would recommend this approach, but let me know your thoughts!")
* Corporate end: formal and structured

### **Energy**

**Slider range:** Calm → Energetic

Sets the pace and enthusiasm of your Double's responses.

* Calm end: measured and composed (e.g. "Sounds good, looking forward to helping!")
* Energetic end: upbeat and enthusiastic

### **Emoji usage**

**Slider range:** None → Heavy

Controls whether and how often your Double uses emojis.

* None: text-only responses
* Heavy: emojis used frequently (e.g. "Great! Let's get started! 🚀")

### **Humor**

**Slider range:** Serious → Playful

Controls how much lightness your Double brings to responses.

* Serious end: professional and matter-of-fact
* Playful end: adds wit and warmth (e.g. "Let's tackle this together! 😄")

### **Topic Flexibility**

Controls how your Double handles off-topic questions.

Use the slider to set the range:

* **Stay on Topic** — your Double strictly sticks to your area of expertise and deflects anything outside it
* **Flexible** — your Double can briefly touch on off-topic questions and relate them back to your expertise (e.g. "I can touch on that briefly. Let me give you a quick answer and relate it back to my expertise.")

#### **What to watch out for**

Setting flexibility too high can cause your Double to drift — answering questions unrelated to you. If your Double is for a specific professional purpose, keep it closer to **Stay on Topic**. Use Custom instructions to add specific guardrails alongside the slider.

### **Auto Suggested Replies**

Toggle this on to automatically surface reply options for visitors during a conversation — keeping the conversation flowing without them having to type from scratch.

Suggested replies appear as clickable prompts after your Double responds, generated dynamically based on the conversation context.

**Tip:** Keep this on for most use cases. It reduces friction for visitors and leads to longer, more productive conversations.

### **Response Structure**

**Person / Point of view** — choose how your Double refers to itself:

* **First person** — "I believe this approach will work well for your project"
* **Third person** — your Double refers to you by name or role (e.g. "Faraz believes this approach will work well for your project")

First person feels more natural for most use cases. Use third person if you want your Double to feel more like a dedicated assistant speaking on your behalf.

### **When Double Doesn't Know Something**

Set how your Double responds when a visitor asks something outside its knowledge.

* **Honest admission** (default) — "That's outside my expertise, but I'd be happy to help with related topics"
* Or describe your own preferred fallback response in the text box on the right

#### **How to write a good fallback**

A good fallback acknowledges the limit, stays on-brand, and redirects the visitor. For example:

* "I don't have that information right now, but feel free to reach out to me directly at \[email]"
* "That's a bit outside my wheelhouse — but happy to help with anything related to \[your expertise]"

Avoid vague fallbacks like "I don't know." They leave visitors with nowhere to go.

**What to know**

* Persona settings apply to the specific Double shown under **Used on** at the top of the page. If you have multiple Doubles, configure Persona separately for each.
* Persona works alongside Knowledge Base — tone and behavior are set here, but what your Double actually knows comes from your knowledge sources.
* Custom instructions override preset behavior when there's a conflict.