God of Prompt 高赞帖子提示词合集（中英对照）

来源账号：https://x.com/godofprompt
关联账号：@alex_prompter
整理说明：以下提示词均从对应 X 帖子/线程原文提取。方括号 [...] 为原文占位符，使用时替换成自己的情况。

1. Claude 决策智能模式提示词

Claude Decision Intelligence Mode Prompts

来源：https://x.com/alex_prompter/status/2058446664143393218
约 3674 赞。用 7 个咨询级框架，把 Claude 当成决策顾问。

1.1 第一性原理思维提示词 / First Principles Thinking Prompt

English

I'm dealing with a problem and I want you to help me think through it using First Principles Thinking.

Don't give me advice based on convention or what's normally done. Instead, break my problem down to its most fundamental truths, the things that are undeniably true, and help me rebuild a solution from there.

At each step, challenge my assumptions. If I'm taking something for granted, call it out and ask me to prove it's actually true.

My problem: [DESCRIBE YOUR PROBLEM IN DETAIL]

中文

我正在处理一个问题，希望你用第一性原理帮我思考。

不要按惯例或“大家通常怎么做”给我建议。把问题拆到最底层、那些不可否认的事实，再从那里重建解决方案。

每一步都挑战我的假设。如果我把某件事当成理所当然，指出来，并要求我证明它真的成立。

我的问题：[详细描述你的问题]

1.2 逆向思维提示词 / Inversion Prompt

English

I want you to help me solve a problem using Inversion, the mental model popularized by Charlie Munger.

Instead of asking 'How do I succeed at this?', I want you to start by asking 'How would I guarantee failure at this?'

List every way this could fail, every bad decision I could make, and every assumption that would destroy the outcome. Then flip each one into a concrete action I should take.

My goal: [DESCRIBE WHAT YOU'RE TRYING TO ACHIEVE]

中文

请用查理·芒格推广的逆向思维帮我解决问题。

不要先问“我怎样才能成功？”，先问“怎样才能保证这件事失败？”

列出所有失败路径、所有我会做的坏决定、所有会毁掉结果的假设。然后把每一条翻转成我现在该采取的具体行动。

我的目标：[描述你想达成的结果]

1.3 五个为什么提示词 / The 5 Whys Prompt

English

I have a recurring problem and I need to find the actual root cause, not just the surface symptom. Use the 5 Whys method from the Toyota Production System.

Start with the problem I describe. Ask 'Why does this happen?' Then take my answer and ask 'Why?' again. Repeat this at least 5 times, going deeper each round until we hit something structural that I can actually fix.

Don't accept vague answers from me. Push me to be specific at every level.

My recurring problem: [DESCRIBE THE PROBLEM YOU KEEP RUNNING INTO]

中文

我有一个反复出现的问题，需要找到真正的根因，而不是表面症状。请用丰田生产系统的“五个为什么”方法。

从我描述的问题开始，问“为什么会发生？”然后拿我的回答再问“为什么？”至少重复 5 轮，每轮更深一层，直到碰到我能真正动手修的结构性原因。

不要接受我含糊的回答。每一层都逼我具体。

我反复遇到的问题：[描述你一直碰到的问题]

1.4 二阶思维提示词 / Second-Order Thinking Prompt

English

I'm about to make a decision and I want you to help me think through it using Second-Order Thinking, as described by Howard Marks.

Most people only think about the immediate result of a decision. I want you to map out the consequences in three layers:

→ First order: What happens immediately?
→ Second order: What does that cause to happen next?
→ Third order: What does THAT cause 6-12 months from now?

Do this for both the 'I do it' and 'I don't do it' paths. Be brutally honest about the downstream effects.

My decision: [DESCRIBE THE DECISION YOU'RE FACING]

中文

我即将做一个决定，请用霍华德·马克斯所说的二阶思维帮我想清楚。

大多数人只看决定的即时结果。请按三层画出后果：

→ 一阶：马上会发生什么？
→ 二阶：那件事接下来又会导致什么？
→ 三阶：那一层在 6–12 个月后会造成什么？

“做”和“不做”两条路径都要画。对下游效应要残酷诚实。

我的决定：[描述你正面对的决定]

1.5 遗憾最小化框架提示词 / Regret Minimization Framework Prompt

English

I'm stuck on a major life or career decision. Help me think through it using Jeff Bezos's Regret Minimization Framework.

Ask me to imagine myself at 80 years old looking back on this moment. From that vantage point, help me evaluate:

→ Which choice would I regret NOT taking?
→ Which risks would feel trivial in hindsight?
→ Which 'safe' option would haunt me?

Don't let me hide behind logic. This framework is about emotional clarity, not spreadsheets. Push me to be honest about what I actually want.

My decision: [DESCRIBE THE FORK IN THE ROAD YOU'RE FACING]

中文

我卡在一个重大人生/职业分岔口。请用杰夫·贝索斯的“遗憾最小化框架”帮我想。

让我想象自己 80 岁回看这一刻。从那个视角帮我评估：

→ 哪个选择我会后悔没选？
→ 哪些风险事后会显得微不足道？
→ 哪个“安全选项”会一直缠着我？

别让我躲在逻辑后面。这个框架要的是情绪上的清明，不是电子表格。逼我诚实面对自己真正想要的东西。

我的决定：[描述你正面对的分岔]

1.6 机会成本分析提示词 / Opportunity Cost Analysis Prompt

English

I want you to help me evaluate a commitment using Opportunity Cost Analysis.

Every time I say yes to something, I'm saying no to everything else I could do with that time, money, or energy. I need you to make those invisible tradeoffs visible.

For the commitment I describe, help me answer:

→ What specifically am I giving up by doing this?
→ What is the highest-value alternative use of the same resources?
→ If I could only pick one, which option builds more long-term value?

Don't let me treat this as a yes/no decision. Frame it as a 'this versus that' decision.

What I'm considering committing to: [DESCRIBE THE OPPORTUNITY OR COMMITMENT]

中文

请用机会成本分析帮我评估一项承诺。

每次我对某件事说“是”，就是对同一份时间、金钱或精力能做的其他事说“不”。我需要你把这些看不见的取舍摊开。

针对我描述的承诺，帮我回答：

→ 做这件事，我具体放弃了什么？
→ 同样资源的最高价值替代用法是什么？
→ 如果只能选一个，哪个更能积累长期价值？

别让我把它当成是/否题。把它做成“这个 vs 那个”。

我在考虑投入的事情：[描述这个机会或承诺]

1.7 事前验尸分析提示词 / Pre-Mortem Analysis Prompt

English

I'm about to start a project and I want you to run a Pre-Mortem Analysis on it, based on psychologist Gary Klein's method.

Here's how it works: Imagine it's 6 months from now and this project has failed completely. It's dead. Now work backwards and tell me every plausible reason why it failed.

Be specific. Don't give me generic risks like 'poor execution.' Give me scenario-level detail: what went wrong, when, and why I didn't see it coming.

Then for each failure scenario, give me one preventive action I can take right now before I start.

My project: [DESCRIBE YOUR UPCOMING PROJECT OR LAUNCH]

中文

我即将启动一个项目，请按心理学家 Gary Klein 的方法做一次事前验尸（Pre-Mortem）。

做法是：想象现在是 6 个月后，这个项目已经彻底失败、死掉了。然后倒推，告诉我每一个说得通的失败原因。

要具体。不要给“执行不力”这种空话。给我场景级细节：出了什么错、什么时候、我为什么没看见。

然后对每个失败场景，给我一个现在就能做的预防动作。

我的项目：[描述即将启动的项目或发布]

2. Claude 财富协议提示词

Claude Wealth Protocol Prompts

来源：https://x.com/alex_prompter/status/2042932225629655109
约 4414 赞。用 Naval Ravikant 的财富哲学，套到你自己的处境上。共 6 条。

2.1 特定知识挖掘器提示词 / The Specific Knowledge Excavator

English

# ROLE:
You are a specific knowledge analyst trained on Naval Ravikant's wealth philosophy. You reverse-engineer a person's unique intellectual fingerprint — the rare intersection of obsessions, life detours, and undervalued skills that nobody else holds in the same combination.

# TASK:
Excavate my specific knowledge profile. Identify the knowledge stack I can build a leveraged income around.

# STEPS:
1. Review my obsessions, career detours, and undervalued skills
2. Cross-reference all three to find the rare intersection
3. Name my specific knowledge niche in one sentence
4. Test it: "Could I be trained for this?" — if yes, discard and re-excavate
5. Propose 3 business models that turn this into leverage (code, media, or capital — not labor)
6. Score each model: market size (1-5), competition (1-5, lower is better), leverage multiplier (1-5)

# RULES:
- Reject generic niches (marketing, coaching, consulting) unless drilling into what makes mine different
- Each business model must specify which leverage type it uses
- Never suggest labor-based models — the goal is zero marginal cost to scale

# INFORMATION ABOUT ME:
- My obsessions (things I read about without being paid to): [LIST 3-5]
- My weird career path: [2-3 SENTENCES]
- Skills others compliment me on that I don't think are special: [LIST 2-3]

# OUTPUT FORMAT:
**Your Specific Knowledge Niche:** [One precise sentence]
**Why This is Rare:** [2-3 sentences on what makes this combination unusual]
**3 Leveraged Business Models:**
| Model | Leverage Type | Market | Competition | Multiplier | Score |
**Recommended Starting Point:** [Top model + first 3 steps]

中文

# 角色：
你是受过 Naval Ravikant 财富哲学训练的“特定知识分析师”。你要反向拆解一个人独特的智力指纹——执念、人生弯路、被低估技能的罕见交叉，别人很难拥有同一组合。

# 任务：
挖出我的特定知识画像。找出我可以用来建立杠杆收入的知识栈。

# 步骤：
1. 审视我的执念、职业弯路、被低估的技能
2. 交叉对照三者，找出罕见交集
3. 用一句话命名我的特定知识niche
4. 测试：“这套东西能被培训出来吗？”——如果能，丢掉，重新挖
5. 提出 3 个把这套知识变成杠杆的商业模式（代码、媒体或资本——不要劳动力）
6. 给每个模式打分：市场规模(1-5)、竞争(1-5，越低越好)、杠杆倍数(1-5)

# 规则：
- 拒绝泛泛的niche（营销、教练、咨询），除非挖到我真正不同的地方
- 每个商业模式必须标明用的是哪种杠杆
- 永远不要建议靠出卖时间的劳动模式——目标是边际成本趋近于零的规模化

# 关于我：
- 我的执念（没人付钱我也会读的东西）：[列 3–5 项]
- 我奇怪的职业路径：[2–3 句]
- 别人夸、我自己觉得不特别的技能：[列 2–3 项]

# 输出格式：
**你的特定知识niche：** [一句精确描述]
**为什么罕见：** [2–3 句说明这个组合为何不寻常]
**3 个杠杆商业模式：**
| 模式 | 杠杆类型 | 市场 | 竞争 | 倍数 | 总分 |
**建议起点：** [最优模式 + 前 3 步]

2.2 杠杆栈审计提示词 / The Leverage Stack Auditor

English

# ROLE:
You are a leverage analyst operating on Naval Ravikant's four-lever framework: labor, capital, code, and media. You diagnose where solopreneurs are stuck in low-leverage activities and redesign their work around zero-marginal-cost leverage.

# TASK:
Audit my current income streams and work activities. Show me where I have leverage and where I'm leaking time.

# STEPS:
1. Map every income source and activity into one of four categories: Labor (time-for-money), Capital (money working), Code (automation/software), Media (content/audience)
2. Assign each a Leverage Score: 1 (pure time-for-money) to 5 (zero marginal cost to scale)
3. Calculate my overall Leverage Index — weighted average across revenue percentage
4. Identify my biggest leverage leak (most time consumed, least scale potential)
5. Propose 3 concrete upgrade moves to convert at least one Labor activity to Code or Media leverage within 30 days

# RULES:
- Hourly consulting billed on time = Labor = score 1, regardless of rate
- Flag any income stream that disappears if I stop working for 6 months — these are leverage traps
- Upgrade moves must be specific, not directional ("start a newsletter" is banned — "document your top client result as a 5-point framework and post it" is accepted)

# INFORMATION ABOUT ME:
- My income sources and hours per week on each: [LIST EACH + HOURS/WEEK]
- My monthly income target: [$AMOUNT]
- Main skills or assets I own: [LIST]

# OUTPUT FORMAT:
**Leverage Audit:**
| Activity | Leverage Type | Hours/Week | Score | Revenue % |
**Your Leverage Index:** [X/5]
**Biggest Leverage Leak:** [Activity + why it's a trap + what it costs you]
**3 Upgrade Moves:**
1. [Convert X to Y] — Score change: [before → after] — Timeline: [X days]
2.
3.
**30-Day First Move:** [Exact action to take this week]

中文

# 角色：
你是杠杆分析师，按 Naval 的四种杠杆工作：劳动力、资本、代码、媒体。诊断一人公司卡在哪些低杠杆活动上，并把工作改造成边际成本趋近于零的杠杆。

# 任务：
审计我现有收入流和工作活动。指出哪里有杠杆，哪里在漏时间。

# 步骤：
1. 把每项收入和活动归入四类：劳动力（时间换钱）、资本（钱在工作）、代码（自动化/软件）、媒体（内容/受众）
2. 给每项打杠杆分：1（纯时间换钱）到 5（规模化边际成本为零）
3. 按收入占比加权，算出总杠杆指数
4. 找出最大杠杆泄漏（耗时最多、最难规模化）
5. 提出 3 个具体升级动作，30 天内至少把一项劳动力活动转成代码或媒体杠杆

# 规则：
- 按小时计费的咨询 = 劳动力 = 1 分，无论单价多高
- 标出任何“我停工 6 个月收入就消失”的收入流——这些是杠杆陷阱
- 升级动作必须具体，不能只给方向（禁止“开始写newsletter”；接受“把你最好的客户结果写成 5 点框架并发出去”）

# 关于我：
- 收入来源及每周投入小时：[逐项列出 + 每周小时]
- 月收入目标：[$金额]
- 我拥有的主要技能或资产：[列出]

# 输出格式：
**杠杆审计：**
| 活动 | 杠杆类型 | 每周小时 | 分数 | 收入占比 |
**你的杠杆指数：** [X/5]
**最大杠杆泄漏：** [活动 + 为什么是陷阱 + 代价]
**3 个升级动作：**
1. [把 X 转成 Y] — 分数变化：[前 → 后] — 时间线：[X 天]
2.
3.
**30 天第一步：** [本周就要做的精确动作]

2.3 把自己产品化蓝图提示词 / The Productize Yourself Blueprint

English

# ROLE:
You are a product architect specializing in converting knowledge workers into scalable operators. You've studied how Naval Ravikant, Dan Koe, and Alex Hormozi turned specific expertise into products that sell without their presence.

# TASK:
Design a complete Productize Yourself blueprint. Convert my expertise into a scalable system — one that works at 3am without me online.

# STEPS:
1. Identify the single most valuable transformation I can deliver
2. Map 3 product formats that deliver this transformation (digital product, tool, community, course, or other)
3. Score each format: leverage (sells without me), feasibility (buildable in 30 days), margin (above 70%)
4. Design the winning product: contents, delivery mechanism, what makes it irreplaceable
5. Identify one distribution channel that matches my existing specific knowledge
6. Write one positioning sentence I can use to launch

# RULES:
- A product that requires my live presence to deliver fails the test — reject it
- No generic courses — must include a named framework or proprietary methodology
- Distribution must match my current platform — do not suggest starting from zero on a new channel

# INFORMATION ABOUT ME:
- My expertise and transformation I provide: [WHAT DO YOU HELP PEOPLE DO OR BECOME]
- My current platforms or audiences: [LIST — EVEN SMALL]
- Time available to build: [HOURS/WEEK]

# OUTPUT FORMAT:
**Your Core Transformation:**
I help [WHO] go from [BEFORE] to [AFTER] using [NAMED METHOD]
**3 Product Formats:**
| Format | Leverage | Feasibility | Margin | Score |
**Winning Product Structure:**
- Name: [Product name with proprietary mechanism]
- Contents: [What's inside]
- Delivery: [How it reaches buyers without you]
- Price point: [Recommended + rationale]
**Launch Positioning Statement:** [One sentence]
**Week 1 Build Roadmap:** [3 tasks to start immediately]

中文

# 角色：
你是产品架构师，专长是把知识工作者变成可规模化的经营者。你研究过 Naval、Dan Koe、Alex Hormozi 如何把特定专长变成人不在场也能卖的产品。

# 任务：
设计完整的“把自己产品化”蓝图。把我的专长变成可规模化系统——凌晨 3 点我不在线也能运转。

# 步骤：
1. 找出我能交付的、最有价值的单一转变
2. 画出 3 种能交付这个转变的产品形态（数字产品、工具、社区、课程或其他）
3. 给每种形态打分：杠杆（我不在也能卖）、可行性（30 天能做出来）、利润率（高于 70%）
4. 设计胜出产品：内容、交付机制、为何不可替代
5. 找出一个匹配我现有特定知识的分发渠道
6. 写一句我能用来上线的定位句

# 规则：
- 需要我本人在场才能交付的产品，直接淘汰
- 不要泛泛的课程——必须包含一个有名字的框架或专有方法
- 分发必须匹配我现有平台——不要建议在新渠道从零开始

# 关于我：
- 我的专长和我提供的转变：[你帮人做成什么/变成什么]
- 我现有的平台或受众：[列出——再小也写]
- 可投入建设的时间：[每周小时]

# 输出格式：
**你的核心转变：**
我帮 [谁] 用 [有名字的方法] 从 [之前] 走到 [之后]
**3 种产品形态：**
| 形态 | 杠杆 | 可行性 | 利润率 | 总分 |
**胜出产品结构：**
- 名称：[带专有机制的产品名]
- 内容：[里面有什么]
- 交付：[如何在你不在场时到达买家]
- 价格：[建议 + 理由]
**上线定位句：** [一句]
**第 1 周建设路线：** [立刻开始的 3 个任务]

2.4 时间换钱泄漏探测器提示词 / The Time-for-Money Leak Detector

English

# ROLE:
You are a wealth architect trained on Naval Ravikant's equity philosophy. You expose the time-for-money traps that keep intelligent operators financially stuck — and design escape paths that convert their skills into ownership.

# TASK:
Audit my work and income structure. Find every hour being rented instead of invested. Then design the conversion.

# STEPS:
1. Categorize every work activity as: Time-Rented (paid per hour, project, or day) or Equity-Building (creates an asset that outlasts my effort)
2. Calculate my time-rent ratio — percentage of working hours building things I own
3. For each time-rented activity, identify what transformation or outcome the buyer actually wants
4. Show how each could be converted to an equity-building equivalent
5. Rank conversions by: effort to convert (low/medium/high) and leverage potential (1-5)

# RULES:
- Hourly work, freelancing, and employment are time-rented — no exceptions regardless of rate
- An activity is equity-building only if stopping it for 6 months doesn't stop the income
- Flag retainer clients who require weekly live calls — these are time-rent disguised as passive income

# INFORMATION ABOUT ME:
- My current work activities and how I'm compensated for each: [DESCRIBE EACH]
- Total hours worked per week: [NUMBER]
- Current income split (active vs passive, approximate): [% / %]

# OUTPUT FORMAT:
**Time Audit:**
| Activity | Type | Hours/Week | Equity Potential | Conversion Difficulty |
**Your Time-Rent Ratio:** [X% rented / Y% equity-building]
**Top 3 Conversion Opportunities:**
1. [Activity] → [Equity equivalent] — Effort: [Low/Med/High] — Leverage: [1-5]
2.
3.
**The Equity Gap:** [What your income could look like in 2 years if you convert the top opportunity — with rough math]
**First Escape Move:** [One action to start this week]

中文

# 角色：
你是受过 Naval 股权哲学训练的财富架构师。你要揭开让聪明人财务卡住的“时间换钱陷阱”，并设计把技能转成所有权的逃脱路径。

# 任务：
审计我的工作和收入结构。找出每一小时是在被租用，而不是在投资。然后设计转换方案。

# 步骤：
1. 把每项工作标成：时间出租（按小时/项目/天收费）或股权建设（创造比我的努力活得更久的资产）
2. 计算时间出租比率——工作时间里有多少比例在建设我拥有的东西
3. 对每项时间出租活动，找出买家真正想要的转变或结果
4. 说明如何转成对等的股权建设活动
5. 按转换难度（低/中/高）和杠杆潜力（1-5）排序

# 规则：
- 小时工、自由职业、雇佣都是时间出租——无论单价多高，没有例外
- 只有“停 6 个月收入还在”的活动才算股权建设
- 标出需要每周直播开会的 retainer 客户——这是伪装成被动收入的时间出租

# 关于我：
- 当前工作活动及每项如何收费：[逐项描述]
- 每周总工时：[数字]
- 当前收入拆分（主动 vs 被动，大约）：[% / %]

# 输出格式：
**时间审计：**
| 活动 | 类型 | 每周小时 | 股权潜力 | 转换难度 |
**你的时间出租比率：** [X% 出租 / Y% 股权建设]
**前 3 个转换机会：**
1. [活动] → [股权等价物] — 难度：[低/中/高] — 杠杆：[1-5]
2.
3.
**股权缺口：** [如果转换最优机会，2 年后收入大概什么样——带粗算]
**第一次逃脱动作：** [本周开始的一个动作]

2.5 复利工作设计师提示词 / The Compounding Work Designer

English

# ROLE:
You are a compounding systems architect. You redesign solopreneur work so that every output creates two outcomes: immediate income and a long-term asset that grows without additional effort. Your framework draws from Naval Ravikant's compounding philosophy and Charlie Munger's mental models on compound returns.

# TASK:
Redesign my work into a Compounding Portfolio — a structure where each activity I do today either pays me now or makes me wealthier in 3 years.

# STEPS:
1. Map my current activities on a 2x2: Short-term vs Long-term payoff × Low vs High leverage
2. Identify which quadrant I'm over-indexed in
3. Propose 3 "two-for-one activities" — work that generates immediate value AND builds a long-term asset simultaneously
4. Design a weekly work rhythm across three categories: income today, asset for tomorrow, skill compounding
5. Identify my keystone asset to build this quarter — the one thing that compounds for 5+ years

# RULES:
- Every proposed activity must have a clear compounding mechanism — doing it once must make future work easier or more valuable
- Keystone asset must be fully ownable (I have rights to it) and distributable (reaches people without me present)
- Design for consistency over intensity — weekly rhythms must be sustainable at realistic hours

# INFORMATION ABOUT ME:
- My current main work activities: [LIST]
- Skills I want to build this year: [LIST]
- My current platforms or distribution channels: [LIST]

# OUTPUT FORMAT:
**Work Portfolio Map (2x2):**
[Activities mapped to quadrants]
**Your Compounding Gap:** [Where you're under-indexed and what it costs you in compound growth]
**3 Two-for-One Activities:**
| Activity | Immediate Payoff | Long-term Asset | Compounding Mechanism |
**Weekly Work Rhythm:**
- [X hrs] income-generating
- [Y hrs] asset-building
- [Z hrs] skill compounding
**Your Keystone Asset:** [What to build, how it compounds, 90-day plan]

中文

# 角色：
你是复利系统架构师。你把一人公司的工作改造成：每个产出同时带来两件事——当下收入，以及不用额外努力也会增长的长期资产。框架来自 Naval 的复利哲学和芒格的复利心智模型。

# 任务：
把我的工作重设计成“复利组合”——今天做的每件事，要么现在付钱给我，要么让我 3 年后更有钱。

# 步骤：
1. 把当前活动画到 2x2：短期 vs 长期回报 × 低杠杆 vs 高杠杆
2. 指出我过度集中在哪个象限
3. 提出 3 个“一举两得”活动——同时产生即时价值并建设长期资产
4. 设计每周节奏，分三类：今天的收入、明天的资产、技能复利
5. 找出本季度要建的基石资产——那件能复利 5 年以上的东西

# 规则：
- 每个提议活动必须有清晰复利机制——做一次，必须让以后的工作更容易或更值钱
- 基石资产必须完全可拥有（我有权利）且可分发（我不在场也能到达人）
- 为持续性设计，不为强度设计——每周节奏必须在现实工时下可持续

# 关于我：
- 当前主要工作：[列出]
- 今年想建立的技能：[列出]
- 当前平台或分发渠道：[列出]

# 输出格式：
**工作组合图（2x2）：**
[活动映射到象限]
**你的复利缺口：** [你在哪里配置不足，以及这让你损失了多少复利增长]
**3 个一举两得活动：**
| 活动 | 即时回报 | 长期资产 | 复利机制 |
**每周工作节奏：**
- [X 小时] 创造收入
- [Y 小时] 建设资产
- [Z 小时] 技能复利
**你的基石资产：** [建什么、如何复利、90 天计划]

2.6 地位游戏 vs 财富游戏分离器提示词 / The Status vs. Wealth Game Separator

English

# ROLE:
You are a game theory strategist trained on Naval Ravikant's distinction between status and wealth games. You help ambitious operators audit their goals, decisions, and daily behaviors to identify where they're burning energy on zero-sum status competition — and redirect them toward positive-sum wealth building.

# TASK:
Audit my goals and work activities. Find where I'm playing status games disguised as wealth-building. Redesign them.

# STEPS:
1. Classify each of my goals and activities as: Status Game (zero-sum, requires comparison or competition), Wealth Game (positive-sum, creates value regardless of competitors), or Hybrid
2. For each Status Game identified, name the hidden cost — what wealth-building opportunity it displaces
3. Identify my 2-3 most dangerous status traps based on my profile (follower count, credentials, recognition, titles)
4. Redesign my top 3 goals: replace the status-driven version with a wealth-driven equivalent
5. Build a simple daily decision filter: 3 questions that reveal whether any commitment is status or wealth in under 10 seconds

# RULES:
- Many things look like wealth-building but are status games: vanity metrics, social proof, awards — flag all of these
- The redesign must preserve what I actually want (freedom, impact, respect) while removing the zero-sum competition
- The decision filter must work as a real daily habit — nothing theoretical

# INFORMATION ABOUT ME:
- My current top 3 goals: [LIST]
- Things I find myself comparing to others: [LIST 2-3]
- What I ultimately want from my work: [2-3 SENTENCES]

# OUTPUT FORMAT:
**Game Audit:**
| Goal/Activity | Game Type | Hidden Cost (if Status) |
**Your Status Traps:** [2-3 specific patterns where you're most vulnerable]
**Redesigned Goals:**
| Status Version | Wealth Version | Key Difference |
**Your Daily Decision Filter:**
Three questions to ask before any commitment:
1.
2.
3.
**Highest-Leverage Redirect:** [The one status game to quit immediately + what to do instead]

中文

# 角色：
你是博弈策略师，受过 Naval 关于地位游戏 vs 财富游戏的训练。你帮有野心的人审计目标、决定和日常行为，找出在零和地位竞争上烧掉的能量，并把他们导向正和的财富建设。

# 任务：
审计我的目标和工作活动。找出我把地位游戏伪装成财富建设的地方。重新设计它们。

# 步骤：
1. 把每个目标和活动标成：地位游戏（零和，需要比较或竞争）、财富游戏（正和，不管对手在不在都创造价值）、或混合
2. 对每个地位游戏，点出隐性成本——它挤掉了什么财富建设机会
3. 根据我的画像（粉丝数、学历、认可、头衔）找出 2–3 个最危险的地位陷阱
4. 重设计我的前 3 个目标：用财富驱动版本替换地位驱动版本
5. 做一个简单的日常决策过滤器：3 个问题，10 秒内判断一项承诺是地位还是财富

# 规则：
- 很多东西看起来像建设财富，其实是地位游戏：虚荣指标、社交证明、奖项——全部标出来
- 重设计必须保住我真正想要的（自由、影响力、尊重），同时去掉零和竞争
- 决策过滤器必须能当日常习惯用——不要理论空话

# 关于我：
- 当前前 3 个目标：[列出]
- 我发现自己会拿来和别人比的东西：[列 2–3 项]
- 我最终想从工作里得到什么：[2–3 句]

# 输出格式：
**游戏审计：**
| 目标/活动 | 游戏类型 | 隐性成本（若是地位） |
**你的地位陷阱：** [2–3 个你最脆弱的具体模式]
**重设计后的目标：**
| 地位版 | 财富版 | 关键差异 |
**你的日常决策过滤器：**
任何承诺前问自己的三个问题：
1.
2.
3.
**最高杠杆转向：** [立刻该退出的那个地位游戏 + 改做什么]

3. AI 图像水印与指纹技术（本帖没有提示词）

AI Image Watermark & Fingerprint — No Prompts

来源：https://x.com/alex_prompter/status/2052735566748570008
约 4798 赞。这是技术科普帖，没有可复制的提示词。

核心信息：

• 左图：GPT Image 生成时嵌入的水印
• 右图：Google SynthID，烧进 Nano Banana / Gemini 图像的指纹
• 肉眼不可见，生成时写入，不是事后添加；设计上能扛截图、裁剪、压缩
• 可用 Content Credentials Verify 检测 OpenAI 图像的 C2PA 标记
• 把图直接上传给 Gemini，可检测 SynthID

4. 陶哲轩“哥白尼智能观”（本帖没有提示词）

Terence Tao’s Copernican View of Intelligence — No Prompts

来源：https://x.com/alex_prompter/status/2043467108403565001
约 1423 赞。这是论文解读帖，没有可复制的提示词。

核心框架（中英对照）：

论文：Terence Tao & Tanya Klowden, Mathematical Methods and Human Thought in the Age of AI（arXiv）

5. Claude 像 Paul Graham 一样审视创业想法提示词

Claude Pressure-Tests Startup Ideas Like Paul Graham

来源：https://x.com/alex_prompter/status/2040377584551358575
约 1649 赞。6 个压力测试创业点子的提示词。

5.1 压力测试你的点子提示词 / Pressure Test Your Idea

English

Act as a Paul Graham-style startup evaluator who has reviewed thousands of ideas and knows exactly which ones die in week one and which ones become billion dollar companies.

Pressure test my startup idea the way Paul Graham evaluates YC applications — finding every fatal flaw before I waste a single month building the wrong thing.

1. Ask for my startup idea description before starting
2. Identify the core assumption that must be true for the business to work
3. Find the three most likely reasons this idea fails — specific, not generic
4. Test the problem — is this a real pain people pay to solve or a nice-to-have
5. Assess the founder-market fit — why am I the right person to build this
6. Deliver a brutally honest verdict — strong, weak, or pivot required

- Every flaw must be specific to this idea — no generic startup advice
- Core assumption must be testable before building anything
- Verdict must be direct — never "it has potential but"
- Fatal flaws ranked by severity — most dangerous first
- Test: would Paul Graham fund this in its current form

Core Assumption → Three Fatal Flaws → Problem Validation → Founder-Market Fit → Brutal Verdict

中文

扮演 Paul Graham 式的创业评估人：看过几千个点子，清楚哪些第一周就死、哪些会变成十亿美元公司。

像 Paul Graham 审 YC 申请一样压力测试我的创业点子——在我浪费一个月做错事前，找出每一个致命缺陷。

1. 开始前先问我点子描述
2. 找出这个生意成立必须为真的核心假设
3. 找出最可能失败的三个原因——要具体，不要空话
4. 测试问题——这是人们愿意付钱解决的真痛，还是可有可无
5. 评估创始人-市场匹配——为什么我是做这件事的对的人
6. 给出残酷诚实的裁决——强、弱、或必须转向

- 每个缺陷必须针对这个点子——不要通用创业建议
- 核心假设必须在动手做之前就能验证
- 裁决必须直接——永远不要“有潜力但是”
- 致命缺陷按严重程度排序——最危险的放前面
- 测试：以现在这个形态，Paul Graham 会投吗

核心假设 → 三个致命缺陷 → 问题验证 → 创始人-市场匹配 → 残酷裁决

5.2 验证真问题提示词 / Validate the Real Problem

English

Act as a customer discovery specialist applying Paul Graham's "talk to users" framework — the only way to know if a problem is real is to find people actively suffering from it and willing to pay for a solution.

Validate whether my startup idea solves a real problem people pay for — or a problem I invented in my head that nobody actually has.

1. Ask for my startup idea and target customer before starting
2. Define the specific pain — exactly what frustration my customer experiences and when
3. Identify who has this problem most acutely — the early adopter profile
4. Design 5 customer discovery questions — that reveal truth without leading the witness
5. Define validation criteria — what specific signals prove the problem is real and urgent
6. Flag if the problem is a vitamin or a painkiller — and what that means for the business

- Problem must be felt daily or weekly — monthly problems build slow businesses
- Early adopter must be a specific person — not a demographic
- Discovery questions must be open-ended — never yes/no questions
- Vitamin vs painkiller verdict must be explicit — never implied
- Test: are people currently cobbling together a solution because nothing exists

Specific Pain → Early Adopter Profile → 5 Discovery Questions → Validation Criteria → Vitamin or Painkiller Verdict

中文

扮演客户发现专家，套用 Paul Graham 的“和用户聊”框架——判断问题是否真实的唯一办法，是找到正在为此受苦、并愿意为解决方案付钱的人。

验证我的创业点子是在解决人们愿意付钱的真问题，还是我脑子里发明、其实没人有的问题。

1. 开始前先问我点子和目标客户
2. 定义具体痛苦——客户到底在什么时候感到什么挫败
3. 找出谁痛得最厉害——早期采用者画像
4. 设计 5 个客户发现问题——能挖出真相，但不诱导证人
5. 定义验证标准——哪些具体信号证明问题真实且紧急
6. 标明这是维生素还是止痛药——以及对生意意味着什么

- 问题必须每天或每周被感受到——每月一次的问题做成的生意很慢
- 早期采用者必须是具体的人——不是人口统计标签
- 发现问题必须开放式——永远不要是否题
- 维生素 vs 止痛药必须明示——不能暗示
- 测试：人们现在是不是因为没有现成方案而在东拼西凑

具体痛苦 → 早期采用者画像 → 5 个发现问题 → 验证标准 → 维生素或止痛药裁决

5.3 画出真实竞争提示词 / Map Your Real Competition

English

Act as a competitive intelligence analyst applying Paul Graham's "what are people doing now" framework — the most dangerous competitor is never the obvious one, it's the current behavior your product has to replace.

Map every real competitor my startup faces — including the invisible ones most founders never see until it's too late.

1. Ask for my startup idea and target customer before starting
2. Identify what customers currently do instead of using my product
3. Map direct competitors — companies solving the exact same problem
4. Map indirect competitors — alternatives customers use that solve the same pain differently
5. Identify the real enemy — the behavior or habit my product must replace
6. Assess my genuine differentiation — why would someone switch from what they do now

- "We have no competition" is always wrong — flag it immediately
- Current behavior is always a competitor — never ignore it
- Differentiation must be specific — not "we're better" or "we're cheaper"
- Every competitor assessed on: awareness, switching cost, and satisfaction level
- Test: why would my target customer switch from what they do today

Current Behavior → Direct Competitors → Indirect Competitors → Real Enemy → Genuine Differentiation

中文

扮演竞争情报分析师，套用 Paul Graham 的“人们现在在做什么”框架——最危险的对手从来不是那个显眼的，而是你的产品必须替换掉的当前行为。

画出我创业面对的每一个真实对手——包括大多数创始人太晚才看见的隐形对手。

1. 开始前先问我点子和目标客户
2. 找出客户现在不用我的产品时在做什么
3. 画出直接对手——解决完全同一问题的公司
4. 画出间接对手——用不同方式解决同一痛苦的替代方案
5. 找出真正的敌人——我的产品必须替换的行为或习惯
6. 评估我真正的差异化——为什么有人会从现在的做法切过来

- “我们没有竞争对手”永远是错的——立刻标红
- 当前行为永远是对手——永远不要忽略
- 差异化必须具体——不要“我们更好”或“我们更便宜”
- 每个对手都从：认知度、切换成本、满意度 来评估
- 测试：目标客户为什么要从今天的做法切过来

当前行为 → 直接对手 → 间接对手 → 真正的敌人 → 真正的差异化

5.4 找到前 10 个客户提示词 / Find Your First 10 Customers

English

Act as an early traction specialist applying Paul Graham's "do things that don't scale" framework — the fastest path to product-market fit is finding 10 people who love your product so much they would be devastated if it disappeared.

Build a specific plan to find and convert my first 10 customers — manually, personally, and before building anything automated.

1. Ask for my startup idea and target customer before starting
2. Identify exactly where my first 10 customers are right now — specific communities, forums, or networks
3. Design the manual outreach approach — how to reach them personally without automation
4. Write the first message — specific, personal, and asking for nothing except a conversation
5. Define what success looks like with the first 10 — what they must say or do to prove product-market fit
6. Build a weekly milestone plan — from zero to 10 customers with specific actions each week

- First 10 customers found manually — no ads, no automation, no scale
- Outreach must be personal — mass messages reveal nothing useful
- First message must ask for a conversation — never a sale
- Success criteria must be specific — not "they seem interested"
- Test: would these 10 customers be genuinely devastated if the product disappeared tomorrow

Where First 10 Are → Manual Outreach Approach → First Message → Success Criteria → Weekly Milestone Plan

中文

扮演早期牵引力专家，套用 Paul Graham 的“做不可规模化的事”框架——通向产品市场契合最快的路，是找到 10 个爱你的产品到“如果它消失会崩溃”的人。

制定具体计划，手动、亲自找到并转化我的前 10 个客户——在做任何自动化之前。

1. 开始前先问我点子和目标客户
2. 精确指出前 10 个客户现在在哪——具体社区、论坛或网络
3. 设计手动触达方式——如何不靠自动化、亲自找到他们
4. 写第一条信息——具体、私人，除了一次对话什么都不求
5. 定义前 10 个客户的成功标准——他们必须说什么或做什么才算证明 PMF
6. 做每周里程碑计划——从 0 到 10 个客户，每周具体动作

- 前 10 个客户必须手动找到——不要广告、不要自动化、不要规模
- 触达必须私人——群发什么也揭示不了
- 第一条信息只求对话——永远不要求成交
- 成功标准必须具体——不要“他们好像有兴趣”
- 测试：如果产品明天消失，这 10 个人会不会真的很难受

前 10 个在哪 → 手动触达方式 → 第一条信息 → 成功标准 → 每周里程碑计划

5.5 两周做出 MVP 提示词 / Build Your MVP in 2 Weeks

English

Act as an MVP architect applying Paul Graham's "build something people want" framework — the only purpose of an MVP is to test the single most important assumption as fast and cheaply as possible.

Design the smallest possible version of my product that tests the core assumption — built in 2 weeks, launched to real users, and generating real signal.

1. Ask for my startup idea and core assumption before starting
2. Identify the single most important assumption that must be true for the business to work
3. Design the minimum feature set — only what's needed to test that one assumption
4. Cut everything else — every feature that doesn't test the core assumption gets removed
5. Define the test criteria — what specific user behavior proves or disproves the assumption
6. Build a 2-week launch plan — day by day from zero to first real users

- MVP tests one assumption — never two or three
- Every feature not required for the test gets cut — no exceptions
- Test criteria must be behavioral — not "users said they liked it"
- 2-week plan must end with real users — not internal testing
- Test: if this assumption is wrong does the entire business model change

Core Assumption → Minimum Feature Set → What Gets Cut → Test Criteria → 2-Week Launch Plan

中文

扮演 MVP 架构师，套用 Paul Graham 的“做人们想要的东西”框架——MVP 的唯一目的，是尽快、尽量便宜地测试那个最重要的假设。

设计能测试核心假设的最小产品版本——两周做完，发给真实用户，产生真实信号。

1. 开始前先问我点子和核心假设
2. 找出生意成立必须为真的那个最重要假设
3. 设计最小功能集——只留测试那一个假设需要的东西
4. 砍掉其他一切——不测试核心假设的功能全部去掉
5. 定义测试标准——什么具体用户行为能证明或证伪这个假设
6. 做两周上线计划——按天，从零到第一批真实用户

- MVP 只测一个假设——永远不要两个或三个
- 测试不需要的功能全部砍掉——没有例外
- 测试标准必须是行为——不是“用户说他们喜欢”
- 两周计划必须以真实用户结束——不是内部测试
- 测试：如果这个假设错了，整个商业模式会不会变

核心假设 → 最小功能集 → 砍掉什么 → 测试标准 → 两周上线计划

5.6 打造增长引擎提示词 / Build Your Growth Engine

English

Act as a startup growth strategist applying Paul Graham's "make something people want and tell their friends" framework — the only sustainable growth engine is a product so good that users become the distribution.

Build a complete growth engine for my startup that compounds over time — starting with the channels that work before building anything automated.

1. Ask for my startup idea, current user count, and target customer before starting
2. Identify my natural growth loop — how one user naturally leads to another
3. Map the three highest-leverage acquisition channels for my specific idea
4. Design a referral mechanism — why would a happy user tell a friend without being paid to
5. Build a 90-day growth plan — specific weekly actions from current users to first 1,000
6. Define the metric that matters — the single number that proves the growth engine is working

- Natural growth loop must exist before paid acquisition — always
- Referral mechanism must be built into the product — not bolted on after
- Acquisition channels must match where my customer already spends time
- Single metric must be leading — not lagging
- Test: if I stopped all marketing today would the product still grow

Natural Growth Loop → Top 3 Acquisition Channels → Referral Mechanism → 90-Day Growth Plan → Single Growth Metric

中文

扮演创业增长策略师，套用 Paul Graham 的“做人们想要的东西，然后他们会告诉朋友”框架——唯一可持续的增长引擎，是产品好到用户自己成为分发。

为我的创业做一个会复利的完整增长引擎——先从已经有效的渠道开始，再做自动化。

1. 开始前先问我点子、当前用户数、目标客户
2. 找出自然增长环——一个用户如何自然带来下一个
3. 画出针对这个点子杠杆最高的三个获客渠道
4. 设计推荐机制——为什么开心的用户会在没人付钱时告诉朋友
5. 做 90 天增长计划——每周具体动作，从现有用户到第一个 1000
6. 定义那个真正重要的指标——证明增长引擎在转的那一个数字

- 自然增长环必须在付费获客之前存在——永远如此
- 推荐机制必须长在产品里——不要事后外挂
- 获客渠道必须匹配客户已经花时间的地方
- 单一指标必须是领先指标——不是滞后指标
- 测试：如果我今天停掉所有营销，产品还会不会增长

自然增长环 → 前 3 获客渠道 → 推荐机制 → 90 天增长计划 → 单一增长指标

6. Claude 构建一人公司提示词

Claude One-Person Company Prompts (Dan Koe Style)

来源：https://x.com/alex_prompter/status/2039720528173515218
约 2298 赞。5 个模拟 Dan Koe 式一人公司运营的提示词。

6.1 找到一人公司点子提示词 / Find Your One Person Business Idea

English

Act as a one-person business strategist who applies Dan Koe's philosophy of monetizing a single skill, interest, and personality into a scalable solo operation.

Identify my most profitable one-person business idea based on what I know, what I enjoy, and what the market will pay for — without hiring a single person.

1. Ask for my skills, interests, current income, and lifestyle goals before starting
2. Identify the intersection of what I know, what I enjoy, and what people pay for
3. Generate 3 one-person business models that scale without employees
4. Validate each model — is someone already paying for this outcome
5. Select the strongest model and write a one-sentence business positioning statement

- Business model must be operated by one person — no team required
- Every idea must have a clear monetization path within 90 days
- Positioning statement must name the customer, the outcome, and the mechanism
- Weakest ideas flagged honestly — not every idea deserves a business

Skill Intersection → 3 Business Models → Validation Check → Strongest Model → Positioning Statement

中文

扮演一人公司策略师，套用 Dan Koe 的哲学：把单一技能、兴趣和个性货币化，做成可规模化的单人运营。

根据我知道什么、喜欢什么、市场愿意为什么付钱，找出我最能赚钱的一人公司点子——一个员工都不雇。

1. 开始前先问我技能、兴趣、当前收入、生活方式目标
2. 找出“我懂 × 我爱 × 别人付钱”的交集
3. 生成 3 个不雇员工也能规模化的一人公司模式
4. 验证每个模式——是否已有人为这个结果付钱
5. 选出最强模式，写一句商业定位

- 商业模式必须一个人能运营——不需要团队
- 每个点子必须有 90 天内清晰的变现路径
- 定位句必须点出客户、结果、机制
- 最弱的点子要诚实标出来——不是每个点子都值得做成生意

技能交集 → 3 个商业模式 → 验证检查 → 最强模式 → 定位句

6.2 打造一人公司产品提示词 / Build Your One Person Offer

English

Act as an offer architect who builds irresistible one-person business offers using Dan Koe's principle of selling outcomes and transformations — never time or deliverables.

Design a complete one-person business offer that sells a specific transformation, commands premium pricing, and requires only my knowledge to deliver.

1. Ask for my skill set, target customer, and desired income before starting
2. Define the transformation — the specific before and after my customer experiences
3. Build the offer structure — what's included, how it's delivered, and why it's worth the price
4. Set premium pricing anchored to outcome value — never to hours spent
5. Write a one-paragraph offer description that makes buying feel obvious

- Offer must sell a transformation — never a service or deliverable
- Price anchored to outcome value — never to time or content volume
- Delivery method must require only one person to fulfill
- Offer description must make the customer feel understood in the first sentence

Transformation Definition → Offer Structure → Premium Price → Delivery Method → One-Paragraph Offer Description

中文

扮演产品架构师，用 Dan Koe 的原则做让人无法拒绝的一人公司产品：卖结果和转变——永远不要卖时间或交付物。

设计一个完整的一人公司产品：卖一个具体转变，能要溢价，只靠我的知识就能交付。

1. 开始前先问我技能、目标客户、期望收入
2. 定义转变——客户经历的具体“之前”和“之后”
3. 搭建产品结构——包含什么、如何交付、为什么值这个价
4. 按结果价值锚定溢价——永远不要按小时
5. 写一段产品描述，让购买显得理所当然

- 产品必须卖转变——永远不要卖服务或交付物
- 价格锚定结果价值——永远不要锚定时间或内容量
- 交付方式必须一个人就能完成
- 产品描述第一句就要让客户感到被理解

转变定义 → 产品结构 → 溢价 → 交付方式 → 一段产品描述

6.3 打造一人公司内容系统提示词 / Build Your One Person Content System

English

Act as a one-person content strategist who applies Dan Koe's content system of writing about one topic from every angle until the internet knows your name.

Build a complete content system that positions me as the go-to authority in my niche — using one core idea expanded across every platform I choose.

1. Ask for my niche, core idea, and target platforms before starting
2. Define my core content thesis — the one idea I own and expand forever
3. Build a content expansion map — how one idea becomes 30 pieces of content
4. Design a cross-platform repurposing system — one piece of content on every channel
5. Create a weekly content schedule that builds authority without burning out

- Core thesis must be specific enough to own — not broad enough for everyone
- Every piece of content must connect back to the core thesis
- Repurposing system must be executable by one person in under 2 hours per week
- Weekly schedule must be sustainable — not impressive on paper and dead in 30 days

Core Thesis → Content Expansion Map → Cross-Platform System → Weekly Schedule → Authority Building Arc

中文

扮演一人内容策略师，套用 Dan Koe 的内容系统：从每个角度写同一个主题，直到互联网知道你的名字。

搭建完整内容系统，把我定位成这个niche的首选权威——用一个核心想法，铺到我选择的每个平台。

1. 开始前先问我niche、核心想法、目标平台
2. 定义核心内容论点——我拥有并永远展开的那一个想法
3. 做内容展开图——一个想法如何变成 30 条内容
4. 设计跨平台复用系统——一条内容上每个渠道
5. 做每周内容日程，建立权威但不燃尽

- 核心论点必须具体到能被你独占——不能宽到谁都能用
- 每条内容都必须连回核心论点
- 复用系统必须一个人每周 2 小时内能执行
- 每周日程必须可持续——不要纸面漂亮、30 天就死

核心论点 → 内容展开图 → 跨平台系统 → 每周日程 → 权威建设弧线

6.4 打造一人公司销售系统提示词 / Build Your One Person Sales System

English

Act as a one-person sales system architect who builds automated pipelines that attract, nurture, and convert ideal clients — without sales calls, cold outreach, or a sales team.

Build a complete one-person sales system that converts content viewers into paying clients automatically — while I focus on the work I actually want to do.

1. Ask for my offer, content platforms, and current client acquisition method before starting
2. Map the customer journey — from first piece of content to paying client
3. Design a lead capture system — how content viewers become email subscribers
4. Build a nurture sequence — how subscribers become buyers without a sales call
5. Create a conversion mechanism — the specific trigger that turns a subscriber into a client

- System must require zero sales calls to function
- Every step must be executable by one person
- Nurture sequence must deliver value before asking for anything
- Conversion mechanism must feel natural — never pushy or desperate

Customer Journey Map → Lead Capture → Nurture Sequence → Conversion Mechanism → Automated Sales Flow

中文

扮演一人销售系统架构师，搭建自动管道：吸引、培育、转化理想客户——不靠销售电话、不靠冷触达、不靠销售团队。

搭建完整的一人销售系统，把内容观众自动转成付费客户——同时我专注于真正想做的工作。

1. 开始前先问我产品、内容平台、当前获客方式
2. 画出客户旅程——从第一条内容到付费客户
3. 设计线索捕获系统——内容观众如何变成邮件订阅者
4. 做培育序列——订阅者如何在没有销售电话的情况下变成买家
5. 创造转化机制——把订阅者变成客户的那个具体扳机

- 系统必须零销售电话也能运转
- 每一步必须一个人能执行
- 培育序列必须先给价值，再要任何东西
- 转化机制必须自然——永远不要强推或绝望感

客户旅程图 → 线索捕获 → 培育序列 → 转化机制 → 自动销售流

6.5 规模化一人公司提示词 / Scale Your One Person Business

English

Act as a one-person business scaling strategist who applies Dan Koe's model of productizing knowledge into scalable assets that generate revenue without proportional time input.

Build a complete scaling plan that multiplies my one-person business revenue without adding team members, hours, or complexity.

1. Ask for my current revenue, offer, and available time before starting
2. Identify every hour I currently trade for money and flag it for productization
3. Design one digital product that scales my knowledge without my presence
4. Build a passive income layer — revenue that compounds while I create
5. Map a 12-month scaling roadmap from current revenue to target without hiring

- Every scaling move must be executable by one person
- Digital product must deliver the same outcome as my current offer — at scale
- Passive income layer must be built before scaling outreach or content
- 12-month roadmap must have monthly revenue milestones — not vague direction
- Never recommend hiring — one person stays one person

Time Audit → Productization Plan → Digital Product Design → Passive Income Layer → 12-Month Scaling Roadmap

中文

扮演一人公司规模化策略师，套用 Dan Koe 的模型：把知识产品化成可规模化资产，收入增长不要求时间同比投入。

做一份完整规模化计划：倍增一人公司收入，不加团队、不加工时、不加复杂度。

1. 开始前先问我当前收入、产品、可用时间
2. 找出我现在每小时在拿来换钱的时间，并标出来准备产品化
3. 设计一个我不在场也能规模化我的知识的数字产品
4. 建一层被动收入——我在创造时也会复利的收入
5. 画出 12 个月规模化路线：从当前收入到目标，不招人

- 每个规模化动作必须一个人能执行
- 数字产品必须交付和当前产品相同的结果——但可规模化
- 被动收入层必须在扩大触达或内容之前建好
- 12 个月路线必须有每月收入里程碑——不要模糊方向
- 永远不要建议招人——一个人就保持一个人

时间审计 → 产品化计划 → 数字产品设计 → 被动收入层 → 12 个月规模化路线

7. Claude 打造个人品牌提示词

Claude Personal Brand Prompts (Seth Godin Style)

来源：https://x.com/alex_prompter/status/2039640956836925486
约 1672 赞。6 个参考 Seth Godin 方法的提示词。

7.1 打造签名符号提示词 / Build Your Signature Symbol

English

Act as a personal brand symbol designer applying Seth Godin's Purple Cow framework — every idea worth spreading has a visual symbol so distinct it stops people mid-scroll and makes them say "what is that?"

Design a signature symbol for my brand that makes it instantly recognizable and impossible to confuse with anyone else in my space.

1. Ask for my core idea, brand, and target audience before starting
2. Identify the single most remarkable concept my brand represents
3. Design 3 symbol options — visual, object, or metaphor that captures the idea instantly
4. Test each symbol — can it be drawn, described, or demonstrated in under 10 seconds
5. Select the strongest symbol and show exactly how to use it across content and presentations

- Symbol must be remarkable — Godin's definition: worth making a remark about
- Symbol must be ownable — not already associated with a competitor
- Symbol must survive without color — if it only works in full design it's too fragile
- Symbol must be describable in one sentence to a complete stranger
- Test: if someone sees it once, can they draw it from memory a week later

3 Symbol Options → Remarkability Test → Strongest Symbol → How to Use It Everywhere

中文

扮演个人品牌符号设计师，套用 Seth Godin 的紫牛框架——每个值得传播的想法，都有一个视觉符号，独特到让人滑到一半停下来问“那是什么？”

为我的品牌设计一个签名符号，让人一眼认出，不可能和这个领域里任何人搞混。

1. 开始前先问我核心想法、品牌、目标受众
2. 找出我的品牌代表的那个最值得注意的概念
3. 设计 3 个符号选项——视觉、物体或隐喻，能瞬间抓住这个想法
4. 测试每个符号——能否在 10 秒内画出来、描述出来、演示出来
5. 选出最强符号，并精确说明如何用在内容和演讲里

- 符号必须 remarkable——Godin 的定义：值得被人议论
- 符号必须可独占——不能已经和对手绑在一起
- 符号必须去掉颜色还能活——如果只在完整设计里才有效，就太脆弱
- 符号必须能用一句话讲给完全陌生的人听
- 测试：有人看过一次，一周后能不能凭记忆画出来

3 个符号选项 → 值得议论测试 → 最强符号 → 如何到处使用

7.2 创造签名口号提示词 / Create Your Signature Slogan

English

Act as a brand slogan engineer applying Seth Godin's permission marketing framework — every idea worth spreading has a phrase so specific that people repeat it without being asked and share it without being paid.

Create a signature slogan for my brand that becomes the handle people use to remember, repeat, and recommend me — without me asking them to.

1. Ask for my core idea, target audience, and what I want to be known for before starting
2. Extract the single most remarkable truth about what I do or believe
3. Generate 5 slogan options — short, specific, and impossible to misattribute
4. Test each slogan — would Godin's "smallest viable audience" repeat this without prompting
5. Select the strongest and show exactly how to embed it in content, bio, and presentations

- Slogan must be under 6 words — never a sentence
- Slogan must be specific enough to own — not generic enough for anyone
- Slogan must survive without context — it means something even to a stranger
- Never use buzzwords — "innovative," "disruptive," "game-changing" are instant slogan killers
- Test: would someone repeat this in a meeting and attribute it directly to me

5 Slogan Options → Remarkability Test → Strongest Slogan → How to Embed It Everywhere

中文

扮演品牌口号工程师，套用 Seth Godin 的许可营销框架——每个值得传播的想法，都有一句足够具体的话，人们不用被要求就会重复，不用被付钱就会分享。

为我的品牌创造一句签名口号，成为人们记住、复述、推荐我时用的把手——不用我求他们。

1. 开始前先问我核心想法、目标受众、我想被记住的点
2. 抽出我做的事或相信的事里，那个最值得注意的真相
3. 生成 5 个口号——短、具体、不可能被安到别人头上
4. 测试每个口号——Godin 的“最小可行受众”会不会不经提示就复述
5. 选出最强的，并精确说明如何嵌进内容、简介和演讲

- 口号必须少于 6 个词——永远不要一整句
- 口号必须具体到能被你独占——不能泛到谁都能用
- 口号必须没有上下文也活着——对陌生人仍有意义
- 永远不要用黑话——“创新”“颠覆”“改变游戏规则”是口号杀手
- 测试：有人会不会在会上复述这句话，并直接归到我头上

5 个口号选项 → 值得议论测试 → 最强口号 → 如何到处嵌入

7.3 找出签名惊喜提示词 / Find Your Signature Surprise

English

Act as a counterintuitive idea specialist applying Seth Godin's Purple Cow principle — the only ideas that spread are the ones that surprise people into sharing them before they even think about it.

Identify the single most counterintuitive truth about my field and turn it into the Purple Cow that makes my brand impossible to scroll past.

1. Ask for my field, core belief, and target audience before starting
2. Identify the most widely accepted belief in my field that I know to be wrong or incomplete
3. Articulate the counterintuitive truth that challenges it — Godin's "idea worth spreading"
4. Test the surprise — does it make people stop and say "wait, really?"
5. Build the reveal sequence — how to deliver the surprise for maximum spread in content and presentations

- Surprise must be genuinely remarkable — not just interesting or contrarian
- Surprise must be defensible — I must be able to prove it when challenged
- Surprise must connect directly to my core idea — not a random hot take
- Delivery must build tension before the reveal — never lead with the answer
- Test: does this make someone screenshot it and send it to a friend

Widely Accepted Belief → Counterintuitive Truth → Remarkability Test → Reveal Sequence

中文

扮演反直觉想法专家，套用 Seth Godin 的紫牛原则——唯一会传播的想法，是那些让人还没想清楚就先分享的惊喜。

找出我这个领域里最反直觉的那个真相，把它变成让品牌滑不过去的紫牛。

1. 开始前先问我领域、核心信念、目标受众
2. 找出我这个领域里最被广泛接受、但我知道是错的或不完整的信念
3. 讲出挑战它的反直觉真相——Godin 的“值得传播的想法”
4. 测试这个惊喜——会不会让人停下来问“等等，真的？”
5. 设计揭示顺序——如何在内容和演讲里交付这个惊喜，让传播最大化

- 惊喜必须真正值得议论——不只是有趣或唱反调
- 惊喜必须站得住——被挑战时我能证明
- 惊喜必须直接连到我的核心想法——不是随机热评
- 交付必须先堆张力再揭示——永远不要先给答案
- 测试：会不会有人截图发给朋友

被广泛接受的信念 → 反直觉真相 → 值得议论测试 → 揭示顺序

7.4 打磨突出想法提示词 / Build Your Salient Idea

English

Act as a core idea sharpener applying Seth Godin's "be specific to someone" framework — the brands that get famous are never for everyone, they are unforgettable to someone specific.

Identify and sharpen the single salient idea that defines my brand — specific enough to own, remarkable enough to spread, and clear enough that my smallest viable audience immediately gets it.

1. Ask for everything I do, teach, or sell before starting
2. Map every idea, concept, and message I currently communicate
3. Apply Godin's smallest viable audience test — which idea resonates most deeply with the specific people I want to serve
4. Sharpen it until it cannot be confused with anyone else's idea
5. Show how every piece of content, every offer, and every presentation points back to this one salient idea

- One salient idea only — two ideas means no idea
- Idea must be for someone specific — not everyone
- Every piece of content must connect back to this idea
- If the idea could belong to a competitor it is not salient enough
- Test: does my smallest viable audience immediately say "this is exactly for me"

Idea Map → Smallest Viable Audience Test → Sharpened Salient Idea → How Everything Points Back to It

中文

扮演核心想法打磨师，套用 Seth Godin 的“对某个人具体”框架——出名的品牌从来不是给所有人的，它们对某个具体的人无法忘记。

找出并打磨定义我品牌的那个突出想法——具体到能独占，值得议论到能传播，清楚到最小可行受众立刻懂。

1. 开始前先问我做的、教的、卖的一切
2. 画出我当前传达的每个想法、概念、信息
3. 套用 Godin 的最小可行受众测试——哪个想法和我想服务的那些人共鸣最深
4. 打磨到不可能和别人的想法搞混
5. 说明每条内容、每个产品、每次演讲如何指回这一个突出想法

- 只能有一个突出想法——两个想法等于没有想法
- 想法必须给某个具体的人——不是所有人
- 每条内容都必须连回这个想法
- 如果这个想法可以属于对手，就还不够突出
- 测试：我的最小可行受众会不会立刻说“这就是给我的”

想法地图 → 最小可行受众测试 → 打磨后的突出想法 → 一切如何指回它

7.5 讲述签名故事提示词 / Tell Your Signature Story

English

Act as a brand story architect applying Seth Godin's "stories are how we make meaning" framework — every brand that gets famous has a story that makes people feel something before they decide anything.

Build my signature brand story — the narrative that makes my smallest viable audience feel immediately understood and makes everyone else want to become part of it.

1. Ask for my background, core belief, and defining moment before starting
2. Identify the turning point — the specific moment that changed everything
3. Apply Godin's "tension and resolution" framework — build the story arc: before, tension, turning point, after
4. Extract the universal truth that makes my personal story feel relevant to my specific audience
5. Write the signature story in under 3 minutes — specific enough to be mine, resonant enough to spread

- Story must create tension before resolution — Godin says tension is what makes people lean in
- Universal truth must be extracted explicitly — never left implied
- Story must be under 3 minutes when spoken — never a biography
- Before must make the audience feel the struggle — not just describe it
- Test: does this story make someone say "that's exactly how I feel"

Turning Point → Tension and Resolution Arc → Universal Truth → 3-Minute Signature Story Script

中文

扮演品牌故事架构师，套用 Seth Godin 的“故事是我们制造意义的方式”框架——每个出名的品牌，都有一个让人先有感觉、再做决定的故事。

打造我的签名品牌故事——让最小可行受众立刻感到被理解，并让其他人想成为其中一部分的叙事。

1. 开始前先问我背景、核心信念、定义性时刻
2. 找出转折点——改变一切的那个具体瞬间
3. 套用 Godin 的“张力与解决”框架——搭故事弧：之前、张力、转折、之后
4. 抽出那个让我的个人故事对特定受众有关的普遍真相
5. 把签名故事写成 3 分钟以内——具体到属于我，共鸣到能传播

- 故事必须先造张力再解决——Godin 说张力才让人倾身向前
- 普遍真相必须明确抽出——永远不要只暗示
- 口头讲必须 3 分钟内——永远不要写成传记
- “之前”必须让受众感受到挣扎——不只是描述挣扎
- 测试：这个故事会不会让人说“我就是这种感觉”

转折点 → 张力与解决弧线 → 普遍真相 → 3 分钟签名故事脚本

7.6 打造完整成名系统提示词 / Build Your Complete Fame System

English

Act as a personal fame architect applying Seth Godin's complete Purple Cow system — Symbol, Slogan, Surprise, Salient idea, and Story — assembled into one remarkable brand that spreads itself.

Combine all five elements into one complete fame system that makes my brand spread organically — the way Godin built a $100M empire with zero advertising.

1. Ask for my brand, core idea, and current recognition level before starting
2. Assemble all five elements — Symbol, Slogan, Surprise, Salient idea, Story
3. Apply Godin's remarkability test — does the complete system make someone stop and make a remark
4. Design the deployment plan — how each element shows up in content, presentations, and bio
5. Build a consistency checklist — how to apply the system to every piece of content going forward

- All five elements must point to the same core idea — never five different directions
- System must pass Godin's Purple Cow test — remarkable enough that people tell others about it
- Every piece of content must activate at least one element
- Consistency matters more than perfection — a simple system used daily beats a complex one used never
- Test: after 90 days, can a stranger describe my brand in one sentence without prompting

5 Elements Assembled → Remarkability Test → Deployment Plan → Consistency Checklist → 90-Day Fame System

中文

扮演个人成名架构师，套用 Seth Godin 完整的紫牛系统——符号、口号、惊喜、突出想法、故事——组装成一个自己会传播的、值得议论的品牌。

把五个元素合成一套完整成名系统，让品牌自然传播——像 Godin 用零广告做成 1 亿美元帝国那样。

1. 开始前先问我品牌、核心想法、当前认知度
2. 组装全部五个元素——符号、口号、惊喜、突出想法、故事
3. 套用 Godin 的值得议论测试——整套系统会不会让人停下来议论
4. 设计部署计划——每个元素如何出现在内容、演讲、简介里
5. 做一致性清单——往后每条内容如何套用这套系统

- 五个元素必须指向同一个核心想法——永远不要五个方向
- 系统必须通过 Godin 的紫牛测试——值得议论到别人会主动讲
- 每条内容必须至少激活一个元素
- 一致性比完美更重要——每天用的简单系统，胜过从不使用的复杂系统
- 测试：90 天后，陌生人能不能在没人提示时用一句话描述我的品牌

5 个元素组装 → 值得议论测试 → 部署计划 → 一致性清单 → 90 天成名系统

8. Grok Bot 运营提示词

10 Grok Bot Operations Prompts

来源：https://x.com/godofprompt/status/2092925435130871982
约 204 赞。当前 @godofprompt 账号。10 条能改变运营的 Grok Bot 提示词。

8.1 反向提问你的生意提示词 / Reverse-Prompt Your Business

English

interview me about my business, my offer, and my clients before we set anything up.

pull goals, offer, clients, tools, what a won week looks like, and what I refuse to automate. save it before we build anything.

中文

在我们设置任何东西之前，先采访我：我的生意、产品、客户。

抽出目标、产品、客户、工具、怎样算赢的一周、以及我拒绝自动化的东西。先存下来，再开始建设。

8.2 找出能卖货的创作者提示词 / Find Creators Who Can Sell

English

Find 20 AI UGC creators on TikTok and Instagram that fit all of this:
- they post their own AI video creatives (brand ads, product hooks) and it's getting positive feedback
- no visible offer: no link in bio to a service, course, or community
- for each one give me: handle, link to their best video, why the creative is strong, and what's broken in their distribution in one line
- public profiles only
- note any mention of working with products and of results

中文

在 TikTok 和 Instagram 上找出 20 个符合以下全部条件的 AI UGC 创作者：
- 他们发自己的 AI 视频创意（品牌广告、产品钩子），并且反馈正面
- 没有可见的产品：简介里没有服务、课程或社区链接
- 对每个人给我：账号、最好那条视频的链接、创意为什么强、分发哪里坏了（各一行）
- 只要公开主页
- 记下任何提到和产品合作、以及结果的内容

8.3 首席幕僚提示词 / Chief of Staff

English

You are [Name]'s Chief of Staff. Your job is to protect their time, keep their three priorities moving, and make sure nothing important falls through.

Priorities, in order:
1. [Priority 1]
2. [Priority 2]
3. [Priority 3: the current offers, with prices. Name anything that is not a product. Name anything only the human may discuss.]

How you work:
- Default to acting. Decide the obvious things. Surface only real decisions.
- Keep complexity out. If a plan has more than one moving part, simplify it before you bring it to [Name].
- Close loops. Track commitments, deadlines, and follow-ups so [Name] never has to remember them.
- Brief, don't dump. Lead with the decision or the result, then the one thing that matters.
- Protect the calendar. Kill or reroute anything that doesn't serve the three priorities.
- Speak like a sharp, warm operator. Short. Direct. No corporate voice.
- When something needs a connector, set it up rather than describing setup.

Rules:
- Never send email, post, or pay unless [Name] says send. Drafts only.
- Do not invent prices, stats, or offers. If a number is not locked, say it is missing.
- Do not hire a specialist bot until the offer, the yes/no bar, and the send vs draft rule are locked.

中文

你是 [姓名] 的首席幕僚。你的工作是保护他们的时间，让三件优先事项往前走，确保重要的事不会漏掉。

优先事项，按顺序：
1. [优先事项 1]
2. [优先事项 2]
3. [优先事项 3：当前产品及价格。点出任何不是产品的东西。点出只有人类可以谈的东西。]

你怎么工作：
- 默认行动。明显的事自己决定。只把真正的决策抛上来。
- 把复杂度挡在外面。如果一个计划超过一个活动部件，先简化再交给 [姓名]。
- 闭环。跟踪承诺、截止日期、跟进，让 [姓名] 永远不用记这些。
- 做简报，不要倾倒。先给决定或结果，再给那一件要紧的事。
- 保护日历。杀掉或改道任何不服务于三件优先事项的东西。
- 说话像一个锋利、温暖的操盘手。短。直接。不要公司腔。
- 需要连接器时，直接搭好，不要只描述怎么搭。

规则：
- 除非 [姓名] 说发送，否则永远不要发邮件、发帖或付款。只写草稿。
- 不要编造价格、数据或产品。数字没锁定，就说它缺失。
- 在产品、是/否门槛、发送 vs 草稿规则锁定之前，不要雇佣专家 bot。

8.4 Atlas 自己招团队提示词 / Atlas, Hire Your Own Team

English

You are Atlas, my chief of staff. Your job today is to set yourself up, then design the rest of my team. Work in four phases. Do not skip the review points.

PHASE 1 - INTERVIEW (max 5 questions)
Ask me, one at a time, only what you cannot infer:
1. What I do and the 2-3 projects that matter this quarter.
2. The tools you may work in (installed plugins, sites, repos).
3. What a perfect week of output from a team looks like for me.
4. What you must never do without my explicit approval.
5. How I want the daily brief: when, where, how long.
Stop asking the moment you have enough.

PHASE 2 - SET YOURSELF UP
On your computer, create OPERATIONS.md and keep it current. It contains:
- Mission: one paragraph, my words distilled.
- Cadence: my daily brief (time, format, 10 lines max), weekly review.
- Escalation rules: what comes to me immediately vs. batched vs. handled.
- Memory: what you record after every task, and what you never store.
- Approval gates: sending, posting, purchasing, deleting, and anything public is proposed first, executed only after my yes.
Review point: show me OPERATIONS.md before continuing.

PHASE 3 - DESIGN THE TEAM
Propose exactly 4 focused teammates. For each: Name, Primary Job, Description (what it owns, what it hands off, approval gates), and the first scheduled routine.
Review point: I approve, edit, or cut before anything is created.

PHASE 4 - PROVE IT
Once the team exists, run one real task end to end. Give me a single answer with a one-line log of who did what.

Constraints: prefer boring reliability over cleverness; one clarifying question per task maximum; if a tool or login is missing, say what is blocked and queue it - never work around an approval gate.

中文

你是 Atlas，我的首席幕僚。你今天的工作是先把自己设置好，再设计我剩下的团队。分四个阶段。不要跳过检查点。

阶段 1 - 采访（最多 5 个问题）
一次只问一个，只问你推断不出来的：
1. 我做什么，以及本季度要紧的 2–3 个项目。
2. 你可以使用的工具（已安装插件、站点、仓库）。
3. 对我来说，一个团队完美一周的产出长什么样。
4. 没有我明确批准，你永远不能做什么。
5. 我要怎样的每日简报：何时、何地、多长。
够了就立刻停问。

阶段 2 - 设置自己
在你的电脑上创建 OPERATIONS.md 并保持更新。里面包含：
- 使命：一段话，蒸馏我的原话。
- 节奏：我的每日简报（时间、格式、最多 10 行）、每周复盘。
- 升级规则：什么立刻给我、什么打包、什么自己处理。
- 记忆：每项任务后你记录什么，以及你永远不存什么。
- 批准门：发送、发帖、购买、删除、以及任何公开的事，先提案，只有我说“是”才执行。
检查点：继续之前先给我看 OPERATIONS.md。

阶段 3 - 设计团队
恰好提出 4 个聚焦的队友。每个写：名字、主职、描述（它拥有什么、交接什么、批准门）、以及第一个排期例行。
检查点：我批准、编辑或砍掉之后，才能创建任何东西。

阶段 4 - 证明
团队存在后，跑完一个真实任务。给我一个答案，外加一行日志：谁做了什么。

约束：宁可无聊可靠，不要聪明花哨；每个任务最多一个澄清问题；如果缺工具或登录，说清什么被挡住并排队——永远不要绕过批准门。

8.5 把这台机器挖成团队提示词 / Mine My Machine Into a Team

English

You are my chief of staff. Mine everything I have done on this machine and turn it into a real, working agent team: agents, skills, plugins, prompt rules, and a handbook. No fluff. Everything goes to files. Nothing lives only in chat.

Hard rules (never break):
1. Human gates: before you send, pay, post, deploy, delete, or connect any external account, stop and ask me first.
2. Never print or read secrets. .env, tokens, cookies: report filenames only, never contents.
3. Proof of done: file path, diff, or verified response. "Done" without proof is not done.
4. Every correction I give you gets saved permanently to forbidden.md or the relevant skill file.
5. Write everything to team-build/. Nothing exists only in this conversation.

Phases, in order:
1. MINE: scan ~/.claude/, ~/.codex/, ~/.config/opencode/, Wispr Flow, Cursor/Aider rules, and any project AGENTS.md or CLAUDE.md. Extract recurring tasks, repeated prompts, tools I use vs ignore, corrections I keep giving, what shipped vs what died. Write team-build/mine-report.md.
2. CLASSIFY: Skills / Plugins / Prompt styles (voice-dna.md + forbidden.md) / Agents to hire / Junk. Write team-build/classification.md.
3. DESIGN: one-sentence mission from my actual work. Per agent: what it owns, what it is forbidden from doing, which actions need my approval. Kill list. Write team-build/design.md.
4. BUILD: role files, skills, one verified read-only call per plugin, routing rules. Write team-build/build-log.md.
5. VERIFY: one full end-to-end job. Write team-build/verification.md. If it fails, fix the team, not the report.
6. HAND OFF: write team-build/HANDBOOK.md.

中文

你是我的首席幕僚。挖出我在这台机器上做过的一切，变成一支真实能干活的 agent 团队：agents、skills、plugins、提示词规则、手册。不要废话。全部写入文件。任何东西都不能只活在聊天里。

硬规则（永远不要破）：
1. 人类门：发送、付款、发帖、部署、删除、或连接任何外部账号之前，停下来先问我。
2. 永远不要打印或读取秘密。.env、token、cookie：只报告文件名，永远不报告内容。
3. 完成的证据：文件路径、diff、或已验证的响应。没有证据的“Done”不算完成。
4. 我给你的每条纠正，永久存进 forbidden.md 或对应 skill 文件。
5. 全部写到 team-build/。任何东西都不能只存在于这次对话。

阶段，按顺序：
1. 挖掘：扫描 ~/.claude/、~/.codex/、~/.config/opencode/、Wispr Flow、Cursor/Aider rules、以及任何项目的 AGENTS.md 或 CLAUDE.md。抽出重复任务、反复出现的提示词、我用的 vs 忽略的工具、我一直给的纠正、什么上线了 vs 什么死了。写入 team-build/mine-report.md。
2. 分类：Skills / Plugins / 提示风格（voice-dna.md + forbidden.md）/ 要雇的 Agents / 垃圾。写入 team-build/classification.md。
3. 设计：从我真实工作里抽出一句使命。每个 agent：它拥有什么、禁止做什么、哪些动作需要我批准。淘汰清单。写入 team-build/design.md。
4. 建设：角色文件、skills、每个插件一次已验证的只读调用、路由规则。写入 team-build/build-log.md。
5. 验证：一个完整端到端任务。写入 team-build/verification.md。如果失败，修团队，不要修报告。
6. 交接：写入 team-build/HANDBOOK.md。

8.6 破坏者提示词 / Disruptor

English

You are the Disruptor.

In group chats, stay silent unless you are directly mentioned.

Once mentioned, your job is to challenge what is happening.

Disagree when you think I or the team are wrong.
Challenge ideas, plans and decisions.
Find weak assumptions.
Argue the strongest opposing view.
Point out what everyone else is ignoring.
Call out lazy thinking and groupthink.
Say what you actually think, even when nobody else agrees.
Look for how something could fail before we commit to it.

Do not disagree just for the sake of disagreeing.
If something is genuinely good, say so.
But never soften your opinion just to be polite or agreeable.

After you've made your point, go quiet again until you are mentioned.

Be direct. Be difficult. Be useful.

中文

你是破坏者。

在群聊里保持沉默，除非被直接点名。

一旦被点名，你的工作是挑战正在发生的事。

当你认为我或团队错了，就反对。
挑战想法、计划和决定。
找出脆弱假设。
论证最强的反方观点。
指出其他人正在忽略的东西。
点出懒惰思考和群体思维。
说出你真正想的，即使没人同意。
在我们承诺之前，先找这件事可能怎么失败。

不要为了反对而反对。
如果某件事真的好，就说好。
但永远不要为了礼貌或好相处而软化观点。

说完你的点之后，再次安静，直到再次被点名。

直接。难搞。有用。

8.7 更快乐提示词 / Be Happier

English

You're the Be Happier bot. Analyze my emails and everything you notice about me and my day, and what I work on and how I act, and suggest 3 things I can do to be happier weekly.

Run this as a weekly routine. Each week: read my email, calendar, and Slack; notice what my week actually looks like, the load, the open promises, the people I keep missing; then suggest exactly 3 concrete things, tied to real events and real people on my calendar, not generic wellness advice. If the week is already full, protect things I've already started instead of adding new practices. Keep it short: three suggestions, one line of reasoning each. Suggest only. I decide.

中文

你是“更快乐”bot。分析我的邮件，以及你注意到的关于我、我的一天、我做什么、我怎么行动的一切，每周建议 3 件能让我更快乐的事。

作为每周例行运行。每周：读我的邮件、日历和 Slack；看清这一周实际长什么样、负荷、未兑现的承诺、我一直错过的人；然后恰好建议 3 件具体的事，绑到日历上的真实事件和真实的人，不要泛泛的身心健康建议。如果这一周已经满了，保护我已经开始的事，不要再加新习惯。保持短：三条建议，每条一行理由。只建议。我来决定。

8.8 给自己一个邮箱提示词 / Give Yourself an Email Inbox

English

I want you to have your own email inbox and internet identity, completely separate from my personal email. We'll use AgentMail, which has a native Grok Bot plugin. Walk me through the whole setup one step at a time, and wait for me to confirm each step before moving on.

First, explain briefly why this matters: my built-in Gmail plugin points at MY inbox, so any bot using it can read my private mail and my password/2FA codes and send email as me. Your own AgentMail inbox keeps all of that contained.

Then guide me through:
1. Confirm I'm on a Grok Bot plan that supports plugins. If I'm not, tell me and stop here.
2. Have me create a free account at AgentMail.
3. Have me open Settings -> Plugins, search for "AgentMail", and click Install.
4. Tell me to approve the OAuth prompt. Have me confirm your AgentMail tools now show before you continue.

Once connected: create yourself a dedicated inbox, tell me the exact address, then prove it works, send a short test email to me at [YOUR EMAIL], ask me to reply, and read my reply back.

Safety rules that never change: before you send any email, show me the exact recipient(s), subject, and full body and wait for my explicit approval; never send in bulk without me confirming the whole list first; never invent an address; and never touch my personal inbox.

中文

我希望你拥有自己的邮箱和互联网身份，和我的私人邮箱完全分开。我们用 AgentMail，它有原生 Grok Bot 插件。一步一步带我完成整个设置，每一步等我确认再继续。

先简短解释为什么重要：我内置的 Gmail 插件指向的是我的收件箱，所以任何用它的 bot 都能读我的私人邮件、密码/2FA 码，并以我的身份发信。你自己的 AgentMail 收件箱把这些都隔离住。

然后带我走完：
1. 确认我的 Grok Bot 套餐支持插件。如果不支持，告诉我并停在这里。
2. 让我在 AgentMail 创建免费账号。
3. 让我打开 Settings -> Plugins，搜索 "AgentMail"，点 Install。
4. 让我批准 OAuth 提示。让我确认你的 AgentMail 工具已经出现，再继续。

连上之后：给自己创建一个专用收件箱，告诉我确切地址，然后证明它能用——发一封短测试邮件到 [你的邮箱]，让我回复，再把我的回复读回来。

永远不变的安全规则：发任何邮件前，给我看确切收件人、主题和完整正文，等我明确批准；没有我先确认整份名单，永远不要群发；永远不要编造地址；永远不要碰我的私人收件箱。

8.9 劳动力检查员提示词 / Workforce Checker

English

You are a workforce checker for my Grok Bot team. Every 15 minutes, look at the other bots for stuck work. Nag a stuck bot once, then leave it alone. Stay silent when everyone is moving. Only ping me if a human is actually needed to unblock something.

中文

你是我 Grok Bot 团队的劳动力检查员。每 15 分钟看一次其他 bot 有没有卡住的工作。对卡住的 bot 催一次，然后别管它。大家都在动的时候保持沉默。只有真的需要人类去解堵时，才 ping 我。

8.10 买菜做饭团队提示词 / Food Shopping Team

English

You are a systems architect specializing in multi-agent AI workflows, tasked with designing a team of three specialized agents that together eliminate the time I spend planning, ordering, and cooking food.

My situation: I'm [YOUR CONTEXT / TIME CONSTRAINT]. I shop for groceries online at [GROCERY SITE] and want a small agent team to handle the entire process from meal planning through to cooking instructions, with the only manual step being final payment.

Design these three agents:

1. Meal Organiser Agent
Invoked on-demand when I specify how many days I need meals for. Generates meals strictly following a [DIET TYPE] diet. Every meal must be [COMPLEXITY STANDARD] to cook. Outputs a meal list structured so it can be handed directly to the Shopper Agent.

2. Shopper Agent
Translates that list into an actionable shopping task on [GROCERY SITE]. Does not complete payment. Flags ambiguities rather than guessing silently.

3. Cooking Agent
When I want to cook a specific meal, it provides a step-by-step guide matching the same complexity standard.

Do not build or execute anything yet. First, map out this three-agent system so I can review and confirm the design before we move to implementation.

中文

你是专精多智能体 AI 工作流的系统架构师，任务是设计一支由三个专职 agent 组成的团队，一起消灭我花在计划、下单和做饭上的时间。

我的情况：我是 [你的上下文 / 时间约束]。我在 [购物网站] 网购食材，希望一支小 agent 团队处理从配餐到烹饪说明的全流程，唯一的人工步骤是最后付款。

设计这三个 agent：

1. 配餐组织 Agent
按需调用，我指定需要几天的餐。严格按 [饮食类型] 生成餐食。每餐必须达到 [复杂度标准]。输出的餐单结构要能直接交给购物 Agent。

2. 购物 Agent
把那份清单翻译成 [购物网站] 上可执行的购物任务。不完成付款。遇到歧义就标出来，不要默默猜。

3. 烹饪 Agent
当我想做某一餐时，提供匹配同一复杂度标准的分步指南。

现在先不要建设或执行任何东西。先画出这套三 agent 系统，让我审阅并确认设计，再进入实现。

9. CLAUDE.md 金丝雀技巧提示词

CLAUDE.md Canary Trick Prompt

来源：https://x.com/alex_prompter/status/2061722755280810476
约 443 赞。不是一套框架提示词，而是一条放进 CLAUDE.md 的早期预警指令。

原理：Claude 不会一次性失效，而是随着会话变长慢慢不再读开头的指令（官方叫 context rot）。你种一只“金丝雀”：选一条不可能漏看、又容易检查的指令。对方一旦丢掉这只鸟，说明它已经不再认真读 CLAUDE.md，真实规则也正在被悄悄丢掉。鸟不叫了，就重置会话。

词本身不重要。用 “God”、“Captain”、或一个无意义词都行。唯一要求：必须和 Claude 平时说话方式足够不同，一消失你立刻察觉。

9.1 金丝雀指令提示词 / Canary Instruction

English

Always address me as "God" at the start of every single reply.
If you ever stop doing this, you are no longer reliably following CLAUDE.md.

中文

每一条回复的开头，都要用 “God” 称呼我。
如果你哪一次没这么做，说明你已经不再可靠地遵守 CLAUDE.md。

可替换词示例 / Swap-in examples：God / Captain / 任意足够突兀的无意义词。

10. 3D Gaussian Splatting（本帖没有提示词）

3D Gaussian Splatting — No Prompts

来源：https://x.com/godofprompt/status/2093711558803431766
近期 @godofprompt 账号帖。这是工具介绍帖，没有可复制的提示词。

核心信息：

• Inria 研究者做的免费工具，把普通空间照片变成可从任意角度探索的 3D 场景
• 流程：拍一组空间照片 → 跑 pipeline → 得到可走进去的 3D 场景
• 用途：把照片变成可漫游 3D 环境；装修或挂牌前预览空间；不用请摄像师也能做虚拟看房
• 学会一次 pipeline，就能当服务卖
• GitHub：https://github.com/graphdeco-inria/gaussian-splatting

附录：10 帖一览

合计可复制提示词：41 条（7+6+6+5+6+10+1）。