为你的 AI Agent
读懂"你是谁"Teach your AI Agent
who you actually are
这套工具的目标不是 self-help——是 AI configuration。
通过七个维度看清自己的思维结构，你才能告诉 AI 该如何与你协作。
Not self-help — this is AI configuration.
See your thinking across seven dimensions, then tell AI how to work with it.
做这件事的两个目标Two goals behind this
认识自己是路径，配置 AI 才是终点。Knowing yourself is the path; configuring AI is the destination.
- 把你的认知模式写进系统提示词，让 AI 不再用通用模板对你Put your patterns into the system prompt — no more generic template
- 明确告诉它你的归因方式、决策启动、典型失误Tell it how you attribute, how you start decisions, where you slip
- 完成全部七个维度后，可一键复制可粘贴的配置文本After all seven, copy a paste-ready config block
- 知道 AI 应该在你的哪些时刻主动介入、问哪种问题Know when AI should step in, and what to ask
- 知道 AI 不该做什么——什么会触发你的防御、消耗判断Know what AI shouldn't do — what triggers your defenses
- 识别自己的盲区，让 AI 在你看不见的地方多看一眼Spot your blind spots, so AI watches where you can't
换句话说——你不是在做心理测试，你在为 AI 写"用户手册"。
但要写得准，你得先读懂手册的主人公。
You're not taking a personality test — you're writing a user manual for your AI.
To write it well, first read the protagonist.
七维度 · 三层认知结构Seven dimensions, three layers
从外层的观察方式，到内层的自我定义，再回到决策的运行层。From reading the outside, to defining yourself, back to how decisions run.
因果心智模型Causal Model
你如何理解原因与结果？为什么有些事成、有些事不成？How do you read cause and effect? Why does one thing work and another doesn't?
概率直觉Probability Intuition
你如何评估可能性？面对不确定时的本能反应是？How do you weigh likelihood? What's your gut reaction to uncertainty?
系统思维Systems Thinking
你能看到几步之外的连锁？反馈回路出现时怎么读？How many steps ahead do you see? How do you read feedback loops?
身份模型Identity Model
你怎么定义自己？这个定义悄悄做了什么？How do you define yourself? What does that quietly do?
叙事模型Narrative Model
过去、现在、未来在你脑里是什么形状？What shape do past, present, future take in your head?
约束模型Constraint Model
你没说出口、但其实一直在用的边界是什么？What lines do you keep without saying out loud?