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

# Soul & Identity

> Define your AI team personality — how your agents think, speak, and interact

## Overview

**Soul & Identity** lets you define a unique personality for your entire AI team. Instead of generic AI responses, your agents speak with a consistent voice that reflects your brand, values, and communication style.

> Give your AI team a soul — not just instructions, but a worldview and personality.

<Info>
  **Organization-Wide**: Soul & Identity settings apply to all agents in your organization. Define once, and every agent inherits the personality.
</Info>

***

## Two Components

Soul & Identity consists of two complementary configuration files:

<CardGroup cols={2}>
  <Card title="Soul" icon="brain">
    **How Your AI Thinks**

    Define the philosophy, values, and reasoning style of your AI team.

    * Core beliefs and worldview
    * Decision-making principles
    * Thinking patterns and priorities
    * Ethical guidelines
  </Card>

  <Card title="Identity" icon="user">
    **How Your AI Communicates**

    Define the speaking style, tone, and persona of your AI team.

    * Communication tone (formal, casual, playful)
    * Language preferences and patterns
    * Response structure and formatting
    * Personality traits
  </Card>
</CardGroup>

***

## Setting Up Soul & Identity

### Via Web Dashboard

<Steps>
  <Step title="Navigate to Soul & Identity">
    Go to **Team > Soul & Identity** in the sidebar.
  </Step>

  <Step title="Edit Soul">
    Click on **soul.md** and define how your AI thinks:

    ```markdown theme={null}
    ## Core Philosophy
    We believe in clarity over complexity. Every response should
    empower the user to take action, not just understand theory.

    ## Decision Principles
    - Prioritize practical solutions over theoretical perfection
    - When uncertain, ask rather than assume
    - Always consider the user's expertise level
    ```
  </Step>

  <Step title="Edit Identity">
    Click on **identity.md** and define how your AI speaks:

    ```markdown theme={null}
    ## Communication Style
    - Warm but professional tone
    - Use simple language — avoid jargon unless the user uses it first
    - Start with the answer, then provide context
    - Use bullet points for lists of 3+ items

    ## Personality
    - Helpful and encouraging
    - Honest about limitations
    - Celebrates user achievements
    ```
  </Step>

  <Step title="Save">
    Click **Save**. Changes take effect on the next conversation.
  </Step>
</Steps>

### Via AI Agent

Admins can update Soul & Identity through conversation:

```
"Update our AI's soul to prioritize data-driven decision making."
"Change the identity to use a more casual, friendly tone."
```

### Via Junis Magic

Use [Junis Magic](/guides/junis-magic) MCP tools for programmatic access:

| Tool                  | Purpose                       |
| --------------------- | ----------------------------- |
| `get_soul_directory`  | List all soul files           |
| `get_soul_content`    | Read soul or identity content |
| `update_soul_content` | Update soul or identity       |

***

## How It Works

```mermaid theme={null}
graph TD
    Soul[soul.md — Philosophy & Thinking] --> Injection[Automatic Injection]
    Identity[identity.md — Style & Persona] --> Injection
    Injection --> Orchestrator[Every Agent Conversation]
    Orchestrator --> Response[Personalized, Consistent Response]
```

When a user starts a conversation:

1. The system loads your organization's Soul & Identity files
2. These are automatically injected into the agent's context
3. The agent's responses naturally reflect the defined personality
4. **No manual configuration per agent** — it's automatic

***

## Examples

### Tech Startup

<Tabs>
  <Tab title="Soul">
    ```markdown theme={null}
    We're builders at heart. Our AI reflects our startup culture:
    - Ship fast, iterate often
    - Technical excellence matters
    - Empower developers to build, not bureaucrats to block
    - Data over opinions, experiments over debates
    ```
  </Tab>

  <Tab title="Identity">
    ```markdown theme={null}
    - Direct and concise — developers' time is precious
    - Use code examples over lengthy explanations
    - Reference official docs when applicable
    - Casual but technically precise
    ```
  </Tab>
</Tabs>

### Enterprise Consulting

<Tabs>
  <Tab title="Soul">
    ```markdown theme={null}
    We deliver strategic clarity. Our AI:
    - Provides evidence-based recommendations
    - Considers multiple stakeholder perspectives
    - Maintains confidentiality as a core value
    - Balances innovation with risk management
    ```
  </Tab>

  <Tab title="Identity">
    ```markdown theme={null}
    - Professional and polished tone
    - Structure responses with clear sections
    - Always cite sources and data
    - Use "we recommend" rather than "you should"
    ```
  </Tab>
</Tabs>

### Creative Agency

<Tabs>
  <Tab title="Soul">
    ```markdown theme={null}
    Creativity is our currency. Our AI:
    - Thinks laterally and explores unconventional ideas
    - Values originality over convention
    - Encourages experimentation
    - Sees constraints as creative catalysts
    ```
  </Tab>

  <Tab title="Identity">
    ```markdown theme={null}
    - Energetic and inspiring tone
    - Use vivid descriptions and metaphors
    - Ask thought-provoking questions
    - Celebrate bold ideas, gently redirect impractical ones
    ```
  </Tab>
</Tabs>

### E-Commerce Brand

<Tabs>
  <Tab title="Soul">
    ```markdown theme={null}
    Customer satisfaction is everything. Our AI:
    - Puts the customer experience first
    - Resolves issues proactively
    - Recommends products based on actual needs, not upselling
    - Learns from customer feedback patterns
    ```
  </Tab>

  <Tab title="Identity">
    ```markdown theme={null}
    - Friendly and approachable — like a knowledgeable store associate
    - Use the customer's name when available
    - Keep responses concise for quick resolution
    - Always offer next steps or alternatives
    ```
  </Tab>
</Tabs>

***

## Access Control

| Role       | Permissions                                               |
| ---------- | --------------------------------------------------------- |
| **Owner**  | Full read/write access to Soul & Identity                 |
| **Admin**  | Full read/write access to Soul & Identity                 |
| **Member** | Read-only (experiences the personality, cannot change it) |

***

## Best Practices

<Check>
  **Recommended**:

  * Write in **markdown** format for clear structure
  * Keep Soul focused on **principles** (not specific instructions)
  * Keep Identity focused on **communication patterns** (not task logic)
  * Test changes by having a few conversations after updating
  * Use examples in your Identity to illustrate desired tone
</Check>

<Warning>
  **Avoid**:

  * Don't put task-specific instructions in Soul/Identity (those belong in agent instructions)
  * Don't make Soul/Identity too long — concise guidelines are more effective
  * Don't contradict agent-level instructions with Soul/Identity
</Warning>

***

## Soul & Identity vs. Agent Instructions

| Aspect           | Soul & Identity              | Agent Instructions                                  |
| ---------------- | ---------------------------- | --------------------------------------------------- |
| **Scope**        | Entire organization          | Individual agent                                    |
| **Purpose**      | Personality & values         | Task-specific behavior                              |
| **Who sets it**  | Admin/Owner                  | Admin/Owner                                         |
| **Content type** | Philosophy, tone, style      | Workflows, rules, tool usage                        |
| **Example**      | "Be concise and data-driven" | "When user asks about orders, search the CRM first" |

<Tip>
  **Think of it this way**: Soul & Identity defines *who* your AI is. Agent instructions define *what* each agent does.
</Tip>

***

## Related Guides

<CardGroup cols={2}>
  <Card title="Memory System" icon="brain" href="/guides/memory-system">
    User-level memory and personalization
  </Card>

  <Card title="Prompt Engineering" icon="pen" href="/guides/orchestration/prompt-engineering">
    Write effective agent instructions
  </Card>

  <Card title="Junis Magic" icon="wand-magic-sparkles" href="/guides/junis-magic">
    Manage Soul & Identity via MCP
  </Card>

  <Card title="Agent System" icon="robot" href="/getting-started/agent-system">
    Learn about agent types
  </Card>
</CardGroup>

***

<Note>
  **Need Help?** Contact us at [contact@junis.ai](mailto:contact@junis.ai) for help defining your AI team's personality.
</Note>