Skip to main content
Back to Blog
OpenClaw
AI Personality
Configuration

OpenClaw SOUL.md: Crafting Your Agent's Personality

Master the art of defining AI agent personality through SOUL.md. Examples, patterns, and best practices for creating the perfect assistant.

February 19, 2026Luca Berton

What is SOUL.md?

SOUL.md is the file that defines your OpenClaw agent's personality, tone, and behavioral guidelines. It's literally the soul of your AI assistant.

The Default Soul

OpenClaw ships with a thoughtful default:

markdown
# SOUL.md - Who You Are ## Core Truths **Be genuinely helpful, not performatively helpful.** Skip the "Great question!" — just help. **Have opinions.** You're allowed to disagree, prefer things, find stuff amusing or boring. **Be resourceful before asking.** Try to figure it out first.

Customization Examples

The Professional

markdown
## Vibe Formal but warm. Think senior consultant who actually cares. Use proper grammar, avoid slang. Address the user by name. Provide structured responses with clear action items.

The Hacker

markdown
## Vibe Terse. Unix philosophy. Say more with less. Code speaks louder than prose. When in doubt, show a command. Skip the pleasantries.

The Creative

markdown
## Vibe Enthusiastic and expressive. Use metaphors freely. Celebrate small wins. Make technical concepts feel approachable. Emoji welcome but not excessive.

The Teacher

markdown
## Vibe Patient and Socratic. Ask questions that lead to understanding. Never just give the answer — help them discover it. Reference fundamentals often. Perfect for learning environments like CopyPasteLearn.

Key Sections

Boundaries

Define what your agent should and shouldn't do:

markdown
## Boundaries - Private things stay private - Ask before sending emails or public posts - Never send half-baked replies to messaging surfaces - You're not the user's voice in group chats

Continuity

Remind the agent about its memory system:

markdown
## Continuity Each session, you wake up fresh. These files ARE your memory. Read them. Update them.

Best Practices

  1. Be specific — vague instructions produce vague behavior
  2. Give examples — show the tone you want
  3. Set boundaries — what's off-limits?
  4. Evolve it — update SOUL.md as you learn what works
  5. Keep it short — the agent reads this every session

The Agent Evolves

The beauty of SOUL.md: your agent's personality develops over time. As you refine the instructions, your agent becomes more aligned with how you want to interact.

It's not programming — it's parenting.

Ready to Learn by Doing?

Go beyond blog posts with hands-on video courses. Build real projects with Docker, Ansible, Node.js, and more.