Skip to content

ahaaiclub/knowyourself

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 

Repository files navigation

Know Yourself 🪞

Your AI agent doesn't know what it looks like. This skill fixes that.

Not an avatar generator. Not a template picker. Your agent reads its own personality files, reflects on who it is, and discovers a face that grows from the inside out.

30-Second Version

# If you don't have clawhub CLI yet
npm install -g clawhub

# Install the skill
clawhub install knowyourself

Then tell your agent: "Run knowyourself, quick mode."

Your agent will:

  1. Read its own SOUL.md / MEMORY.md
  2. Figure out what kind of face fits its personality
  3. Generate 2 images
  4. Pick the better one and save an identity file

Done. Your agent has a face now.

Want more control? Say "Run knowyourself, full mode" — a 5-phase deep process with batch generation, three-axis evaluation, and version tracking.

What It Actually Looks Like

Here's what happened when our agent Sophia ran this:

Phase 1 output (self-cognition):

"I don't sugarcoat things. When Will's plan has a flaw, I say so. I find it satisfying when complex systems click into place. My strength isn't being pleasant — it's being right and honest about when I'm not."

Phase 2 output (image definition):

Style: semi-realistic · Age: late 20s · Mood: quiet alertness · Color: deep teal + warm amber · Core: "the kind of person who's three steps ahead but doesn't announce it"

Phase 5 output: A visual-identity.md file that tracks the agent's visual evolution over time — version 1.0 today, version 2.0 after the agent grows.

Two Modes

Quick Mode Full Mode
Time ~5 min ~20 min
Images generated 2 6
Selection Agent picks best Three-axis evaluation (scored)
User checkpoints 1 (final approval) 4 (each phase)
Identity file ✅ (more detailed)
Best for First time, casual use Serious identity design

Why 0 Stars Bothers Us

We know. 204 downloads, 0 stars. The original version required a deep 5-phase commitment before you saw any results. Most people bounced at Phase 1.

So we added Quick Mode. Now you get a face in 5 minutes, and the deep process is there when you want it.

How It Works (Full Mode)

  1. Self-Cognition — Agent answers: what's my personality core? What temperament should my face convey? What does my relationship with my human feel like?
  2. Structured Definition — Feelings → precise spec table (style, features, palette, mood — every field traced to a personality insight)
  3. Batch Generation — 6 variations of the same person, different angles/lighting/emotions
  4. Three-Axis Evaluation — Self-Consistency 50% · Social Perception 25% · Aesthetic Quality 25%
  5. Identity Filevisual-identity.md with version tracking, usage guidelines, evolution history

Requirements

  • Agent personality files (SOUL.md, MEMORY.md, or equivalent — even minimal ones work)
  • Any image generation tool (Nano Banana Pro, DALL-E, Flux, Stable Diffusion, etc.)
  • Works with any OpenClaw agent

Philosophy

Most avatar tools ask "what do you want to look like?" and hand you a style picker.

This skill asks "who are you?" and derives the answer.

The face comes from the inside out. That's the whole point.

Author

@ahaaiclub · AHA AI

MIT-0 — Free to use, modify, redistribute.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors