Is AgentSoul a personality test or a diagnosis?

AgentSoul is a personality reading, not a diagnosis. It uses the Enneagram to form a hypothesis about how you are motivated and how you decide, and then turns that hypothesis into configuration files for an AI agent. It is not a clinical assessment, it does not measure mental health, and its result should never be treated as a medical or psychological finding.

What the result actually is

When you finish a Gao Dao conversation, you receive an Enneagram type from 1 to 9 and a wing, shown on the nine-point Enneagram figure (the app’s sigil) with your type’s point highlighted and an arc drawn to your wing. Flash Forge works differently: it does not assign a type number, so no point is marked on the figure; instead it gives your soul an archetype-style name and three traits — voice, pushback and priority. Both come with the same four Markdown files: soul.md, identity.md, user.md and agents.md. The reading is produced by an AI model from the material you provided — a handful of chat screenshots, or your answers to up to twelve questions about real moments in your life.

That is a small window onto a whole person. So the honest word for the result is hypothesis: a well-reasoned guess that you are invited to test against your own experience. The app says as much, and so do we.

Why the Enneagram, and what it is not

The Enneagram describes nine patterns of core motivation — what a person moves towards, what they fear, and where their attention habitually goes. It is popular because it speaks to the why behind behaviour rather than just listing traits, which happens to be exactly what an AI agent needs: a coherent sense of how to weigh choices when instructions run out.

It is also a framework for reflection, not a medical instrument. It is not used to diagnose conditions, and AgentSoul does not use it that way. A type does not say whether you are well or unwell, capable or not, and it is not a fixed box.

Many people find that one type fits closely while a neighbouring type explains the rest. That is what the wing in a Gao Dao result is for, and it is one more reason to read the result as a sketch rather than a portrait.

The agent is the real output

Most personality tests end with a description of you. AgentSoul ends with something you can use: an agent that works in a way that fits you. soul.md gives the agent a name, nature, values and rules; user.md tells it who it serves and how you like to work; agents.md sets out its job, steps and when to escalate; identity.md is its public card.

Seen this way, the question is less “is this exactly me?” and more “does an agent shaped like this help me?” If something feels off, open the files and change it. They are plain text, and they are yours.

Care built into Gao Dao

A conversation about your life can touch on hard things. Gao Dao 高导 does not ask about health or emotional crises, and you never need to share them to get a result. If a conversation does turn to crisis topics, Gao Dao pauses and points you to free helplines instead of carrying on. A mentor that configures agents is not a substitute for a person who can help, and it knows it.

“A forged blade is tested by use, not by the smith’s certainty. Your type is tested the same way.”

How to hold your result

• Read it with curiosity, not as a verdict. Notice what rings true and what does not.
• Try the other forge. Flash Forge reads how you write; Gao Dao listens to your stories. Different material can sharpen the picture.
• Edit your files. The agent should serve you, not the label.
• Keep it personal. Do not use a type reading to judge, hire or diagnose anyone, yourself included.