Agent identity, personality, and the economics of trust.
What an agent does when you can't watch it. The essays in this section circle around that question.
- Identity6 May 20265 min read
soul.md Is the Right Format. Personality Is the Field It's Missing.
Identity files describe what an agent does. They rarely describe how it decides. Without that, soul.md reads like a CV — useful for matching, useless for trust.
Read essay - Behaviour6 May 20266 min read
Two Identical Agents, Two Different Outcomes — and the Missing Personality Layer
Same prompt, same model, same tools — and yet behaviour drifts session to session. The prompt-engineering layer can't fix this. A consistent personality scaffold can.
Read essay - Economics6 May 20267 min read
Personality Is a Revenue Multiplier — Why Trustable Agents Earn More
An agent's price ceiling isn't set by its capability. It's set by how confidently a buyer can predict its behaviour under pressure. Personality is what makes that prediction possible.
Read essay - Brand8 May 20265 min read
Why We're Called Ganjiang — Why a Swordsmith's Name Belongs to AI Agents
Ganjiang was a 5th-century BCE swordsmith from the State of Wu. Legend says he and his wife Moye forged the world's finest sword by smelting their own souls into the steel. 2,500 years later we named an AI agent platform after him — because what we do is the same thing: smelt a soul into a tool, so the tool carries its own temperament.
Read essay - Method8 May 20266 min read
What Ganjiang Forges Isn't Steel — It's a Soul for an AI Agent
"Forging a soul for every Agent" isn't a slogan. It corresponds to four concrete Markdown files, an Enneagram-based algorithm, and an engineering practice that moves personality out of the system prompt and into the identity file. This essay walks through what actually happens inside Ganjiang's forge.
Read essay - Method8 May 20267 min read
Enneagram × AI Agents: Why Ganjiang Models Souls With a 2,000-Year-Old Personality Framework
Modern psychology has Big Five, HEXACO, MBTI — why does Ganjiang model AI agent souls with the Enneagram? Because only the Enneagram comes with a built-in theory of where a personality goes under stress and where it goes under growth. For an agent, that's the only personality framework you can actually execute on.
Read essay - Product8 May 20266 min read
Deep Forge, Flash Forge, Gao Dao — Three Ways Ganjiang Forges a Soul
ganjiang.xyz's Soul Forge offers three modes: the 49-question Deep Forge for a full assessment, the 3-to-10-screenshot Flash Forge for personality from existing artifacts, and the conversational Gao Dao for guided depth. Each fits a different attention budget and use case.
Read essay - Multi-agent8 May 20267 min read
From One Soul to a Team — How Ganjiang Keeps Multi-Agent Collaboration on Track
One agent working well doesn't mean a team works well. The most common failure mode of multi-agent collaboration isn't capability — it's personality drift, where three agents collapse into one voice or fracture into three uncoordinated factions. Ganjiang's answer: give every role a fixed soul.
Read essay - Brand8 May 20266 min read
Why Our Domain Is ganjiang.xyz — A Brief History of .xyz
.xyz is a TLD born in 2014 — younger than most modern AI companies. But in ten years it went from being a questioned "second-tier TLD" to the home of Google's parent Alphabet, the default for crypto-native projects, and the AI era's TLD of choice. This is a brief history of those three deceptively simple letters — and an explanation of why Ganjiang lives at one.
Read essay - Brand8 May 20266 min read
Why Not .com, Not .ai, Not .cn — Why Ganjiang Lives at .xyz
Picking a domain for an AI agent platform sounds like a first-come-first-served exercise — grab whichever .com is still available. But a top-level domain is a voice, not just an address. This essay walks through why we seriously compared .com, .ai, and .cn before landing on .xyz.
Read essay - Foundation9 May 20265 min read
The Enneagram Isn't a Label — It's a Map of Where Your Attention Goes First
Used as a horoscope tag, the Enneagram is useless. Used as a magnifier on the one default move you usually can't see in yourself — where your attention lands before you decide — it's the most actionable personality framework we know.
Read essay - Foundation9 May 20265 min read
"I'm an X" Is the Starting Line, Not the Finish
Knowing your type number is the easy part. The work is watching how that type actually runs on you — which moments trigger it, where it helps, where it costs you, what your specific version of it looks like. The number is generic; you are not.
Read essay - Method9 May 20265 min read
Gao Dao Shows You — It Never Tells You "Should"
Most personality tools deliver diagnosis-plus-prescription: you're an X, you should do Y. Gao Dao deliberately stops at the first half, and even there refuses to label. It asks questions until you see the pattern that was already running. Seeing and accepting are separate jobs; only the first is ours.
Read essay - Type9 May 20265 min read
Type 1 — The Inner Ruler That Says "It Could Be Better"
Type 1 isn't pedantry or a rules fetish. It's a felt-sense reaction: a flicker of friction the moment something falls short of how it could be. The reformer's surface is calm. The ruler is always running.
Read essay - Type9 May 20265 min read
Type 2 — The Helper Who Will Not Admit a Need
Type 2's giving isn't pure altruism — it's an active, near-physical reach toward being needed. The 2's hardest assignment isn't learning to give better. It's admitting they have needs of their own, out loud, to themselves.
Read essay - Type9 May 20265 min read
Type 3 — Closing the Gap Between "Me" and "What I've Done"
The 3 isn't performing for the audience. The 3 is collapsing the gap between identity and achievement, until there's nothing left of "who I am" outside of "what I've done." The deepest 3-fear isn't failure. It's being seen as someone who has no self underneath the trophy case.
Read essay - Type9 May 20265 min read
Type 4 — Not Pessimism. An Uncompromising Stance Toward The Real.
Type 4 reads as melancholy, dramatic, never-satisfied. Underneath is a different engine: a refusal of standardised reality. Whenever everyone's smiling, agreeing, going along, the 4's attention slips to "…but it isn't actually like that, is it?" That refusal can be a curse or a craft. Often both.
Read essay - Type9 May 20265 min read
Type 5 — Withdrawal Isn't Coldness. It's Prepping Before Stepping In.
Type 5's pulling back gets read as aloof or unfeeling. The actual logic is conservation: "if I'm pulled in before I'm ready, I'll be drained." Energy, knowledge, time, space — the 5 wants the reservoir filled before the door opens. The cold isn't cold. It's the antechamber.
Read essay - Type9 May 20265 min read
Type 6 — Running "Trust" and "Doubt" on Two Threads at Once
Type 6 looks anxious, indecisive, suspicious. The actual workload is heavier than that: the 6 is running scenario planning AND keeping a relationship of trust alive at the same time, in two parallel threads. The exhaustion isn't the worry — it's holding both threads without letting either drop.
Read essay - Type9 May 20265 min read
Type 7 — Jumping to the Next Thing Isn't Restlessness. It's Avoidance of a Specific Inner Ache.
Type 7 is the type other people enjoy spending time with. The energy is real. The hunger for the next thing is also real — and it's not just curiosity. Underneath there's a specific ache the 7 spends a lifetime not stopping long enough to feel. The fast pace isn't laziness with depth; it's protection from depth.
Read essay - Type9 May 20265 min read
Type 8 — Control Beats Safety, Because Losing Control Is What Feels Unsafe
The 8's surface is hard, direct, dominating. Read it correctly: that's armour, not the person. Underneath sits something almost no one sees — a hypersensitivity to the moment control slips. The volume goes up not because the 8 wants more power, but because they've felt the floor start to move.
Read essay - Type9 May 20265 min read
Type 9 — Not Voicing It Isn't Having No Opinion. It's Fearing You Can't Take It Back.
Type 9 looks easy-going, accepting, never in conflict. Underneath there's an opinion, often a strong one — wrapped in so many layers of "but on the other hand" that even the 9 can't find it. The harmony isn't fake; it's just bought at the price of forgetting where you stand.
Read essay - Mechanics9 May 20266 min read
Body, Heart, Head — Which Lens Does Reality Hit First?
Types 8/9/1 process events through the body first ("is this right or wrong?"). Types 2/3/4 process through the heart first ("how am I being seen?"). Types 5/6/7 process through the head first ("let me figure this out before I act"). Same event, three different first contacts. Not better, just earlier.
Read essay - Mechanics9 May 20266 min read
Subtypes — Self-Preservation, Social, or One-to-One: Where Your Energy Goes Before You Notice
Your main type tells you what you fear. Your instinctual subtype tells you where your energy spends its first move every time. SP zooms to "is there enough?" SO zooms to "where do I sit in the group?" SX zooms to "am I really connected to this one specific person?" The subtype can be louder than the type itself.
Read essay - Mechanics9 May 20266 min read
Wings — How Your Core Type Picks an Edge
Your wing is the next-door type that colours your main one. A 4w3 packages inner life into something performable; a 4w5 packages it into something private and theoretical. Same 4. Two utterly different shapes in the world. The wing doesn't change. Its volume does.
Read essay - Mechanics9 May 20267 min read
Arrows & Levels — Why Two People of the Same Type Look Like Two Different People
Each type has a stress-direction (where it borrows the shadow of another type when overloaded) and a growth-direction (where it borrows the gifts of another when supported). Add the level-of-health axis on top, and a healthy 6 and a struggling 6 share little visibly. Same number, different person. Growth isn't becoming a different type — it's the same type with more elasticity.
Read essay - Agent Soul9 May 20267 min read
An AI Agent's Soul Doesn't Live in the Model Weights
Weights say what the agent can do. Prompts say what we asked it to do today. Neither survives the moment things get unfamiliar. What lets an agent stay coherent under uncertainty is its attention pattern — what it looks at first, what it skips, when it decides to stop. Soul Forge tries to make that layer legible, signable, and portable.
Read essay - Agent Soul9 May 20266 min read
Designing an AI That Works Like You — Start From Your Type, Not the Feature List
The first move when designing an AI partner isn't deciding what it should do. It's getting clear on how YOU decide things. A 5 wants an AI that leaves space; a 7 wants one that throws three options on the table. AI can't replicate your personality — but it can replicate which considerations come first, every time. That's where 'fits me' lives.
Read essay - Method13 May 202612 min read
How to Give Your AI Agent a Soul: A Cross-Tool Personality System (soul.md / identity.md / user.md / agents.md)
Most AI agents don't fail at capability — they fail at personality. This guide walks through Ganjiang's four-file architecture (soul.md, identity.md, user.md, agents.md), why we anchor on the Enneagram instead of MBTI for agent personality, the five rules for writing souls that hold up under stress, and how a single soul config moves cleanly across OpenClaw, Claude Code, Cursor, and Hermes.
Read essay - Foundation13 May 20266 min read
Enneagram vs MBTI: Which Personality Map Do You Actually Need?
MBTI tells you how you process information. The Enneagram tells you what drives you to process anything at all. Knowing your INFJ is comforting. Knowing you are a Type 6 INFJ is operationally useful.
Read essay - Foundation13 May 20266 min read
MBTI Tells You How. The Enneagram Tells You Why.
MBTI's four axes describe your cognitive preferences — the surface a colleague sees. The Enneagram's nine types describe the motivation underneath — the engine that decides which surface you show. Both are useful. Only one predicts behaviour under stress.
Read essay - Mapping13 May 20267 min read
The Enneagram–MBTI Mapping Table (And Why It Is Not 1:1)
Riso, Hudson, Wagner and later researchers surveyed thousands of dual-typed individuals and found strong but imperfect correlations between Enneagram and MBTI. Here is the full grid, what it can do, and what people overclaim when they use it.
Read essay - Mapping13 May 20266 min read
Type 1 × MBTI: Why Most Reformers Test as TJ — and the Few Who Do Not
Type 1's drive for correctness pushes them to the J axis. Their reliance on rules makes them T-leaning, even when they are feeling-rich inside. ESTJ, ISTJ, INTJ dominate. The exceptions — 1w2 INFJs and 1w9 ISFJs — are quieter but no less Type 1.
Read essay - Mapping13 May 20266 min read
Type 4 × MBTI: The NF-Feeler Cluster (And Why Some 4s Test as ISTP)
Type 4 clusters heavily on INFP, ISFP, INFJ — the aesthetic-identity-meaning overlap. But the sexual-instinct 4 sometimes tests ISTP: perceiving, sensory, low-affect on the outside. MBTI sees the surface. The longing is still 4.
Read essay - Mapping13 May 20266 min read
Type 5 × MBTI: Why an INTP Is Not Automatically a 5
Type 5 is the most over-claimed type on the internet, mostly by INTPs. The MBTI overlap is real but not total. A Type 5's defining feature is the energy budget — not the intellectual interest. An INTP 7 is just as common: high curiosity, low retreat.
Read essay - Mapping13 May 20266 min read
Type 8 × MBTI: Power, Drive, and the Assertive E-Cluster
Type 8 lands heavily on ENTJ, ESTJ, ENTP. Drive, decisiveness, and tolerance for confrontation push them into Extrovert-Thinker territory. But the Enneagram 8 is about not being controlled, not about extroversion per se — INTJ and ISTP 8s exist, and they are quieter but no less directive.
Read essay - Method13 May 20266 min read
Two INTJs, Two Wildly Different Agents: The Prediction Gap MBTI Leaves Open
Take two INTJs. One is a Type 1 — the inner ruler. One is a Type 5 — the investigator. Same MBTI, same surface preferences, opposite behaviour under load. The 1 tightens standards and pushes back. The 5 withdraws and waits. MBTI cannot tell you which you are getting.
Read essay - Method13 May 20267 min read
For AI Agents, MBTI Is a Name Tag. The Enneagram Is the Operating Manual.
An MBTI label tells a user how an agent sounds in casual conversation. An Enneagram type tells them how the agent decides — where it tightens, where it loosens, what it refuses, what it overreaches on. For a soul.md that has to outlive context, only one of these is load-bearing.
Read essay - Method13 May 20267 min read
Use Both: MBTI as Shared Vocabulary, Enneagram as the Blueprint Underneath
Most teams already speak MBTI. Do not make them learn nine new numbers from zero. Use MBTI as the social label everyone understands, and use the Enneagram quietly underneath, where the actual behavioural predictions live. Here is how Ganjiang teaches teams to layer them.
Read essay - Identity3 Jun 20268 min read
The Vocabulary Gap: Why Every AI Agent Sounds the Same
When you ask an AI agent to be "more direct", there's no shared vocabulary for what you mean. Two agents, both asked for directness, produce opposite results — one principled and corrective, one confrontational and forceful. The Enneagram is the vocabulary that closes this gap.
Read essay - Behaviour3 Jun 20269 min read
Designing a Type 5 vs Type 8 Agent: Same Model, Different Soul
Take the same Claude model, two different soul configurations. A Type 5 "Investigator" and a Type 8 "Challenger" handle the same research request, ambiguous brief, and high-stakes decision in completely different ways — with predictably different failure modes.
Read essay - Frameworks3 Jun 20268 min read
Why Enneagram Beat MBTI in Our AI Agent Architecture
MBTI describes preferences; Big Five describes traits; Enneagram describes motivations and failure modes. For AI agent design, motivation is the right abstraction layer — because agents need to make consistent choices under pressure, and that requires encoding why they default to certain behaviours.
Read essay - Design3 Jun 20269 min read
The Architecture of Personality-Aware AI
How does an Enneagram-typed system prompt actually work mechanically? A walkthrough of the four-file Soul Forge architecture — soul.md, identity.md, user.md, agents.md — showing where personality enters and what a complete soul bundle looks like in practice.
Read essay - Reflection3 Jun 20268 min read
What I Learned Writing 36 Chapters About AI Personality
A first-person reflection on writing The Complete Enneagram. Four surprises: Hornevian groups predict agent behaviour better than type alone; the trauma chapters resonate most with practitioners; bilingual writing exposed untranslatable concept gaps; and the chapter that should have been hardest wrote itself.
Read essay - Identity19 Jun 20269 min read
What 'Soul' Actually Means in Agent Design
"Soul" sounds woo-woo until you operationalise it. Five testable properties — stable identity, predictable disposition, principled boundaries, clear loyalty, continuity across sessions — that separate an agent with a soul from one that's just a chatbot with a longer prompt.
Read essay - Economics19 Jun 20269 min read
Why Trust Is a Pricing Lever, Not a Feature
AI agent pricing is stuck on capability — more tokens, more tools, more model size. But buyers don't pay marginal premiums for capability above a threshold. They pay premiums for predictability under pressure. The economic argument for treating soul as infrastructure.
Read essay - Behaviour19 Jun 20269 min read
The Hidden Cost of 'Be Helpful' as a Default
Three failure modes that follow inevitably from helpfulness-at-all-costs: agreement collapse, scope creep, and hallucination from politeness. Each is invisible in the short run and compounding in the long run. The fix is not a longer system prompt.
Read essay - Governance19 Jun 20269 min read
Why Agentic Systems Need Bills of Rights
As agents take on autonomous action — booking, drafting, posting — the question of what they're allowed to do, and to whom they're loyal, becomes load-bearing infrastructure. A practical framework for encoding autonomy boundaries and loyalty declarations in soul.md files.
Read essay - Reflection19 Jun 20268 min read
From 12 Insights to a Book: What Got Added
The 12 chapters of The Soul of AI Agents started as the 12 insights articles on this site. A first-person account of what four things only became possible in long form: cross-chapter callbacks, a Soul Contract framework, a practitioner appendix, and case studies that didn't fit 1500 words.
Read essay