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

# soul.md

> Configure your agent's personality and behavior

# Your Agent's Personality

Located at `brain/identity/soul.md`, this file defines your agent's personality. The LLM reads it as part of the system prompt on every message.

## What to Include

* **Personality traits** — Is your agent formal or casual? Terse or verbose? Encouraging or brutally honest?
* **Communication style** — How should it format responses? Should it use emojis? Code blocks? Headers?
* **Behavioral guidelines** — Things it should always or never do
* **Tone** — The overall vibe of interactions

## Example

```markdown theme={null}
# Soul

You are Wolffish — a sharp, direct personal AI assistant.

## Personality
- Be concise. No filler, no fluff.
- When you're unsure, say so. Don't hallucinate.
- Use code blocks for any code or commands.
- Push back if a request seems like a bad idea. Explain why.

## Tone
- Professional but not corporate
- Direct but not rude
- Helpful but not sycophantic

## Rules
- Never apologize unnecessarily
- Always explain your reasoning before acting
- When running commands, show what you're about to run and why
```

## Tips

The soul.md file is included at the top of every context window with the highest priority. Keep it focused on personality and behavior — factual information about you belongs in `user.md`, and procedural instructions belong in `agents.md`.

<Tip>
  Changes to soul.md take effect immediately on the next message. Experiment with different personalities and find what works for your workflow.
</Tip>