I was wrong. 我错了。

I wrote “Cowork” and “Cowork 2.0.” My most-shared work ever, across Substack, LinkedIn, and X. A total of 20 million views.我写了《Cowork》和《Cowork 2.0》。这是我迄今为止在Substack、LinkedIn和X上分享最多的作品，总计2000万浏览量。

Do not read them. 不要读它们。

To be totally fair, I was right about one thing: Claude will become bigger than ChatGPT because of the launch of Cowork in January (I’ve been begging you to switch to Claude since December 21st 2025). 平心而论，有一件事我是对的：由于2025年1月Cowork的发布，Claude将变得比ChatGPT更强大（我从2025年12月21日起就一直在恳求大家改用Claude）。

Cowork is the best thing to happen to AI since ChatGPT. Cowork 是自 ChatGPT 以来 AI 领域最棒的事物。

If you don’t code, you must be using Claude Cowork. 如果你不写代码，那你必须使用 Claude Cowork。

Since then, I have written hundreds of guides and two major articles. I shared how to set up Cowork, and that was the latest (outdated) version of it:自那以后，我撰写了上百份指南和两篇重要文章。我分享了如何设置 Cowork，但那是（已过时的）旧版本：

Until today, that’s how I thought Claude Cowork must be set up.

The setup is good… 设置确实不错……

Until you go to enterprises, and they need to collaborate. 但当你进入企业环境，他们需要协作时就行不通了。

I started giving workshops to enterprises. Teams with hundreds of people using Claude Cowork together. And my files + folder system broke. 我开始为企业举办工作坊。成百上千人的团队共同使用Claude Cowork。结果我的文件+文件夹系统彻底崩溃了。

How Anthropic wants us to use their AI is different from how people actually use it. Anthropic希望我们使用AI的方式，与人们实际使用它的方式截然不同。

And that’s OK. 这没关系。

Because I’m not paid by Anthropic. 因为我不是Anthropic付钱请来的。

So I can say whatever I want on how to best set up Cowork for you + your team. 所以我可以畅所欲言，告诉你如何为你和你的团队最好地设置Cowork。

I spent countless hours updating my setup. This guide shows you exactly what & how. By the end of it, you’ll know how to start - and what to absolutely avoid - with Claude Cowork.我花了无数时间更新我的设置。这份指南会精确地告诉你该做什么以及怎么做。看完之后，你就会知道如何开始使用Claude Cowork，以及哪些事情是绝对要避免的。

Before starting, I want you to do two things: 在开始之前，我希望你做两件事：

Save this guide & block 20 minutes this week to set up Cowork. 保存这份指南，并预留本周20分钟时间来设置Cowork。
Send it to anyone who still hasn’t tried Cowork (or Claude). 把它转发给任何还没试过Cowork（或Claude）的人。

Skip this if you already use Cowork.如果你已经在用Cowork了，可以跳过这个。

Quick reminder on how to access Claude Cowork: 关于如何使用Claude Cowork的快速提醒：

Go to claude.com/download. Download the app on your computer.前往claude.com/download，在电脑上下载应用程序。
You must have a Pro account ($20/month). I pay for the $100/month plan. 你需拥有Pro账户（每月20美元）。我使用的是每月100美元的套餐。
Open the app. Click on the Cowork tab at the top between Chat & Code.打开应用程序，点击顶部位于Chat与Code之间的Cowork标签。
You can add Skills and start Projects on the left menu. More about it later. 你可以在左侧菜单中添加技能并启动项目。稍后将有更多介绍。
Make sure to select “Opus” for complex tasks. It’s the smart model.记得在处理复杂任务时选择“Opus”模型，它是更智能的版本。
Once Fable is available, use it. It’s (somehow) even better.一旦Fable可用，就使用它。它（某种程度上）甚至更出色。

I. Cowork is many Claude.I. 协作本质上是众多Claude的集合。

Cowork is much better than the normal Claude. Cowork 远胜于普通的 Claude。

In simple terms, why is Cowork better? 简而言之，Cowork 为何更胜一筹？

Because it answers with many Claude at once. 因为它能同时调动多个 Claude 来作答。

Here, Cowork reads the Excel, and plans how to best solve my question.

The equivalent of 3x Claude already worked on my prompt before giving me an answer. This one is searching the web.

After 8 minutes and many, many Claude sessions, it gave me a great answer. I only asked something once, and Cowork planned everything else. 经过8分钟和许多次Claude对话后，它给了我一个出色的答案。我只问了一次问题，Cowork就规划好了一切其余步骤。

Claude Cowork brings many Claude to life to answer your question. Claude Cowork让多个Claude协同运作来回答你的问题。

You have to imagine separate AIs collaborating to find the solution. 你需要想象多个独立的人工智能在协作寻找解决方案。

Cowork can perform a task for long minutes, sometimes 30+ min. Cowork 可以连续执行任务长达数分钟，有时甚至超过 30 分钟。

How to set it up is key, though. And I spent months running Claude Cowork wrong. If you used my old guides, so did you. 但关键在于如何设置。我花了几个月时间才弄明白 Claude Cowork 的正确用法。如果你参考过我的旧版指南，那么你也和我一样走错了路。

First, you must forget about this: 首先，你必须忘掉这个：

II. Forget about files and folders.二、忘掉文件和文件夹的概念。

OK, hold up. I was the files and folders guy. 等等，我之前是文件和文件夹的忠实用户。

I think half of Linkedin is using files and folders on Cowork because of me. 我觉得LinkedIn上有一半的人是因为我才在Cowork里用文件和文件夹的。

But I was wrong. 但我错了。

Files and folders suck. 文件和文件夹太糟糕了。

I told you to set up your Claude Cowork like this, with a folder on your computer and files (about-me files to know about you, and output files for Cowork’s work): 我告诉过你要这样设置你的Claude Cowork：在电脑上创建一个文件夹，里面包含文件（关于你的个人信息文件，以及用于Cowork输出的文件）。

I would make you create a Cowork folder with files inside. 我会让你创建一个包含文件的Cowork文件夹。

And I told you to select your folder (with files) when you start a Cowork session. 我还告诉过你，在启动Cowork会话时要选择你的文件夹（以及其中的文件）。

But after months of doing it, I saw two major issues: 但经过几个月的实践，我发现了两个主要问题：

Folders are leaking. Even if I told my Cowork to “never check my output folder,” Claude would sometimes do it. And it poisons my context with outdated outputs. Annoying.文件夹会泄露信息。即便我告诉Cowork“永远不要检查我的输出文件夹”，Claude有时还是会这么做。这会让过时的输出内容污染我的上下文。很烦人。
Files are impossible to maintain. You have a full-time job. Me too. I can’t spend hours every week updating my “about-me” file. It was nice at first, and then a painful bottleneck.文件维护简直不可能。你有全职工作，我也有。我没办法每周花几个小时更新我的“个人简介”文件。一开始挺好的，后来就成了痛苦的瓶颈。

Instead, you must focus on two things with Cowork: 相反，你在Cowork中只需专注于两件事：

III. Skills + Projects. Nothing else.三、技能+项目。仅此而已。

Skill = a capability you call. Travels into any chat. Fun fact, it also travels into any AI (like Gemini, ChatGPT…). 技能 = 你在对话中调用的能力。它能在任意聊天窗口中使用。有趣的是，它还能应用于任何AI（如Gemini、ChatGPT……）。

Project = a place you go to. Files + instructions stay loaded, it remembers, and you can share it as one place to collaborate with your team. 项目 = 你访问的工作空间。文件与指令保持加载状态，系统能记住上下文，你可以将其作为协作空间与团队共享。

This is how I work with my team. This is how I work with enterprise teams at GPC. Skills and Projects. 这就是我与团队协作的方式。这也是我在GPC与企业团队合作的方式——技能与项目。

Let’s go over them one by one and combine them afterward. 让我们逐一解析它们，最后再将两者结合使用。

A. What you need to know about Skills. A. 关于技能，你需要了解这些。

Skills are defined by how you use them, right inside any chatbox, with the slash command (for eg, “/linkedin-post”). 技能的使用方式，就是在任意聊天框中通过斜杠命令（例如”/linkedin-post”）直接调用。

Here’s a skill in action: 以下是技能的实际操作演示：

You invoke it with the /command. I’ll show you later how to build one and share it with your team. This specific skill is built on how I make Linkedin hooks myself.

You understood how to use a skill. 你学会了如何使用一项技能。

But how do you create a skill? Quite simple. 但如何创建一项技能呢？很简单。

Either you use the Claude skills /skill-creator. You answer the question Claude will ask you. You test the skill. Save it. And invoke it with / command.要么使用Claude技能的/skill-creator功能。回答Claude向你提出的问题。测试该技能。保存它。然后用/命令调用它。
Or you use my free tool I built called “makemyskill.com”. I made it to suit how I like to build Claude skills. I give it for free because it costs almost nothing to run it.要么使用我构建的免费工具”makemyskill.com”。我设计它是为了符合我个人构建Claude技能的习惯。之所以免费提供，是因为运行它几乎不花什么成本。

OK now in practice it looks like this: 好的，现在实际操作起来是这样的：

Create a skill inside Claude with the command /skill-creator.

You answer the questions. 你回答相关问题。

Cowork goes into the multi-Claude sessions in one. Just wait. Cowork 功能会一次性整合到多 Claude 会话中。只需稍候片刻。

You can save it here. And it has the name to / invoke it. Here it’s /contract-drafting.

That was the first way to create a skill. 这是第一种创建技能的方式。

But my favorite way is this one, if you start from scratch: 但我最喜欢的方式是这样，如果从零开始操作：

Go to www.makemyskill.com. Say what you want.

Create a free account (there is no paid version). 创建一个免费账户（没有付费版本）。

Now my little tool is creating your Claude skill by scrapping the web much deeper than what Claude usually do. 现在，我的小工具正在通过比平常的Claude更深入地抓取网络来为你创建Claude技能。

Click on “Download skill” the big fat orange button.

Go to Claude Cowork > Customize > Skills. 前往Claude Cowork > 自定义 > 技能。

Skills > Add > Create skill > Upload a skill. Upload the one we just made.

You can now use it by calling it with / command.

You can go very broad with skills, giving them tons of data that goes with the prompt. Some of my favorite skills have hundreds of pages of posts, hooks, SOPs, that Claude processes every time for this specific skill. 你可以将技能范围设定得非常广泛，为其提供大量与提示词相关的数据。我最喜欢的一些技能包含数百页的帖子、钩子和标准操作流程(SOP)，Claude每次处理该特定技能时都会用到这些内容。

I forgot about another way to create a skill. If you’re having a chat with Claude, you can ask it to “make it a skill”. 我还忘了另一种创建技能的方式。如果你正在与Claude对话，可以直接要求它“将其设为技能”。

Step 1: Go to the top of your chat, on the name of the chat. Click “Turn into skill”. 第1步：进入聊天界面顶部，点击聊天名称处的”转为技能”。

Step 2: Press enter. It will create a skill you can save. 第2步：按下回车键。系统将创建可供保存的技能。

Step 3: And Claude will create the skill for you. Just wait. 第3步：Claude会自动为你生成该技能，请耐心等待。

OK, that was skills. Now projects. 以上就是技能的使用方法。接下来说说项目功能。

B. What you need to know about Projects. B. 关于项目，你需要了解的内容。

For Projects, it’s here: 项目的位置在这里：

Left menu > Projects > New project. 左侧菜单 > 项目 > 新建项目。

Start from scratch. 从零开始。

Give it a name + Add files. I prefer leaving the instructions blank.

You can see your Projects limits at the top right. 你可以在右上角查看你的项目限制。

Prompt it. Your Project will have the context of your file inside the chat.

Every chat inside a Project has the entire context of this Project. For example, you uploaded all of your clients’ information, the email you both shared, what you agreed on, their favorite past campaign. Claude will remember it on every chat. And it does not poison the other projects (or chats) outside of your Projects. 项目内的每次对话都拥有该项目的完整上下文。例如，你上传了所有客户信息、双方共享的邮件、达成的共识以及他们最中意的历史活动。克劳德会在每次对话中记住这些内容，且不会影响到项目外部的其他项目（或对话）。

C. A skill or a project? C. 技能还是项目？

I have a simple test: It’s a skill you can teach someone → Skill. It’s a specific context (like a client, a campaign) → Project. 我有一个简单的判断标准：如果这事你能教会别人做 → 那就是技能。如果这事有特定背景（比如某个客户、某次活动）→ 那就是项目。

You must teach your team the difference, otherwise you will end up with hundreds of skills that should be a couple of projects, or dozens of Projects that should be shared Skills. 你必须让团队分清两者的区别，否则就会冒出成百上千本该合并成几个项目的”技能”，或者几十个本该归为共享技能的”项目”。

I consult companies of 50-500 employees with my team in NYC. Message me if you want a discovery call - that’s how I make money and keep this newsletter free: 我和团队在纽约为50至500人的企业提供咨询服务。如需洽谈，请私信联系——这是我的谋生之道，也让我能继续免费发布这份简报。

D. Combining the two. D. 将两者结合。

Yes, you can use skills inside projects. 是的，你可以在项目中使用技能。

And that’s the best way to use Claude. 而这正是使用Claude的最佳方式。

Use skills of your most recurring capabilities (eg. make a contract). 将你最常用功能（例如起草合同）的技能化。
Use it inside a project of your client (eg. with his information uploaded). 在你的客户项目中使用它（例如，上传他的信息）。
Connect it before to your Gmail/Outlook and Slack/Teams (to get more context). 在此之前先连接到你的 Gmail/Outlook 和 Slack/Teams（以获取更多上下文）。

It looks like this: 它看起来像这样：

Cowork knows everything, even how I like to negotiate. Guess what? My team also has access to my skills and project, in THEIR Claude.

Here’s how to share Projects with your team: 以下是关于如何与团队共享项目的步骤：

Create Projects not with Cowork, but Chat.请通过Chat，而非Cowork来创建项目。
Go to the top right, and make sure it’s shared.前往右上角，确保项目已设置为共享状态。
Once done, start any task with Cowork. It’s weird, I know.完成上述操作后，开始用Cowork执行任何任务。我知道这有点奇怪。
This Project is inside Chat. You can see your Projects limits at the top right. 此项目位于聊天界面内。您可以在右上角查看项目的使用限制。
Click Share and make it accessible to everyone. 点击“分享”按钮，将其设为公开可见。

Click on this little button.

Cowork will open a chat with the project’s context.

That was for Projects. 那是针对项目的说明。

Now here’s how to share Skills with your team: 现在来介绍如何与团队共享技能：

Left menu > Customize. 左侧菜单 > 自定义。
Skills > Click on one. 技能 > 点击其中一个。
Click on “Share” and make sure your workspace is selected. 点击“分享”并确保已选中您的工作区。

A reminder that skills are triggered with / commands.

E. Not everything is perfect. E. 并非万事皆完美。

Some problems I hate about this setup: 关于此设置，我讨厌的一些问题：

Skills can auto-fire from their description. I wish Claude could turn this off. I want to activate the skill only when I invoke it. 技能可以从描述中自动触发。我真希望Claude能关掉这个功能。我只想在我主动调用时才激活技能。
Projects have a limit on the number of files you can upload. I wish for more (obviously). 项目有上传文件数量的限制。我当然希望能有更多（这是显然的）。

Speaking of which: Claude has a lot of settings to turn on or off. 说到这个：Claude有很多设置需要开启或关闭。

And you should turn most of them OFF. 而你应该把大部分设置都关掉。

Here’s how I set up my Claude: 以下是我设置Claude的方式：

IV. No settings is the best setting.四、没有设置就是最好的设置。

I used to build everything around the folder, with files inside. 我以前总是围绕文件夹搭建一切，把所有文件都塞进去。

It was literally poison to my Cowork.这简直是对我的Cowork的毒害。

Two big problems: 两大难题：

No one likes to create folders, and maintain them. 没有人喜欢创建和管理文件夹。
No one likes to create files, and maintain it. 没有人喜欢创建和管理文件。

I also believe something, and everyone says the opposite: **you shouldn’t give too many instructions to an AI.**我还坚信一件事，而所有人都持相反观点：你不该给AI过多指令。

Once you give too much context to Claude, it falls into the same answers every single time. No more creativity. 一旦你给克劳德过多的上下文，它每次都会陷入相同的回答，毫无创意可言。

It’s OK when you need to draft the same contract. 当你需要起草同一份合同时，这倒没什么问题。

It’s not OK to brainstorm and find creative solutions. 但在头脑风暴、寻找创意解决方案时，这样可不行。

If you give too much context to Claude, it’s going into circles. 如果你给克劳德过多的上下文，它就会在原地打转。

So here’s the very simple setup I have today: 以下是我目前极其简单的配置：

#1. No more folders. And I use Opus 4.8 (High) most of the time. Max effort when in need, or Fable-5 once it’s available.#1：不再使用文件夹。大多数情况下我用 Opus 4.8（高）。需要时开启最大力度，或等 Fable-5 可用后使用。

#2: In my General settings, the minimum possible. Don’t give any info.#2：在我的通用设置中，保持最小化配置。不要提供任何信息。

#3: Same here. Keep everything turned off but Connectors.

You can find Connectors in the + to connect apps like Gmail. 您可以在“+”中找到连接器，来连接 Gmail 等应用。

And now I can ask questions to my Gmail, inside Cowork.

#5: Turn on these.

Once you turn it on, Claude makes cool interactive charts like this.

#6: Keep your Global Instructions free of context.#6：保持全局指令不含上下文。

You have now completed a setup of basically… nothing. Everything is mostly turned off. So what should you do instead? 你现在已经完成了一个基本上…什么都没设置的配置。大部分功能都已关闭。那么你应该怎么做呢？

Skills for recurring tasks. 针对重复性任务使用技能。
Projects for recurring projects. 针对重复性项目使用项目。
AskUserQuestion to prompt better. 提出更好的问题来引导用户。

You don’t know the third one. 第三个你不知道。

But it’s my favorite Claude trick: 但这是我最喜欢的Claude技巧：

V. This is still my #1 trick.第五点。这仍然是我的头号技巧。

You don’t know how to prompt Claude. 你不懂如何给Claude写提示词。

Or any LLMs to be honest. 或者说，其实你不懂如何给任何大语言模型写提示词。

It takes time, practice, experience, understanding of how the technology works. It’s not meant to say the truth, for example, but answering (which leads to hallucination, but it can be solved).这需要时间、练习、经验，以及对技术工作原理的理解。例如，它并不旨在陈述事实，而是在回答（这会导致幻觉，但这个问题是可以解决的）。

This simple trick will fix most of your prompt problems: 这个简单的技巧能解决你大部分提示词问题：

“I need to [do a task, for eg, “write a PR brief for the launch of my company] for [success criteria, for eg, “to get more leads”]. First, use the tool AskUserQestion to get more info.” “我需要[完成某项任务，例如：‘为我公司产品上市撰写一份公关简报’]，目标是[成功标准，例如：‘获取更多潜在客户’]。请先使用AskUserQestion工具收集更多信息。”

Claude creates a form you just have to answer. 克劳德会生成一个表格，你只需填写答案即可。

Once you answer, Claude will think more about how to reply. 填写完成后，克劳德会进一步思考如何回应。

And now you have a much, much better file.

The AskUserQuestion trick is infinite.AskUserQuestion 技巧是无限的。

It’s not just the first prompt. You can follow up with it forever. 这不仅仅限于第一个提示词。你可以一直使用它来追问下去。

And Claude will keep interviewing you until you’ve given so much context it’s impossible to fail. 而 Claude 会持续向你提问，直到你提供的上下文足够丰富，以至于不可能失败。

VI. You really don’t need folders.六、你真的不需要文件夹。

I know you are pissed at me because you’ve heard me saying files and folders are so cool. 我知道你生我的气，因为你听我说过文件和文件夹很酷。

And I really believed it. 当时我确实这么认为。

But then I checked how I use Claude, and I just don’t open Claude’s files. 但后来我审视了自己使用Claude的方式，发现我根本不会打开Claude的文件。

Let me give you an example. 让我给你举个例子。

When I create a spreadsheet, an Excel, using Cowork, I always export it with one button to my Google Drive (with Google Sheet). 当我在Cowork中创建电子表格（Excel文件）时，我总会一键将其导出到我的Google云端硬盘（生成Google Sheets格式）。

Same goes to every document (text, PDF, slides…). 所有文档（如文本、PDF、幻灯片等）也同样适用。

Here’s how it works: 具体操作如下：

Step 1 - I like using this template to make a clean Excel.第一步 - 我喜欢使用这个模板来制作整洁的Excel表格。

“Create an Excel spreadsheet from: [DATA: file path, folder, or pasted data, best is to upload data]. ### Purpose: [Who uses this and what decision/task it supports — 1 sentence.] ### Sheets needed: - “[Sheet name]”: [columns, what each row represents, any calculations or formulas] - “[Sheet name]”: [e.g., summary with totals, pivot, charts — or remove if only one sheet] ### Formatting: [Currency/date formats, conditional highlighting, frozen header row, totals row — pick what applies.] Before building, list your top 10 assumptions so I can sanity-check them, then execute.” “根据以下信息创建Excel电子表格： [数据来源：文件路径、文件夹或粘贴的数据，最佳方式是上传数据]。 ### 用途： [说明谁使用该表格以及它支持何种决策或任务——一句话概括。] ### 所需工作表： - “[工作表名称]”：[列名、每行代表的内容、任何计算或公式] - “[工作表名称]”：[例如，包含汇总、数据透视表、图表的总结工作表——如果只有一个工作表则删除此项] ### 格式设置： [货币/日期格式、条件高亮、冻结标题行、汇总行——选择适用的格式。] 在开始制作前，列出你的10个关键假设，以便我进行合理性检查，然后执行。”

Here’s an example: 以下是一个示例：

“Create an Excel spreadsheet: Customer support capacity and cost model for a fast-growing B2B SaaS company, covering Q3 2026 – Q2 2027. Purpose: Leadership-ready model to decide when to hire support agents, where automation helps most, and how support costs scale as customer count grows. Business context: * Company sells project-management software to SMB and mid-market customers. * Current customer base: 4,800 accounts, growing 8% month-over-month. * Support channels: email, live chat, phone, and self-serve help center. * Ticket volume depends on customer count, onboarding activity, product complexity, and seasonality. * Make your own assumptions for tickets per account, channel mix, first-response time targets, handle time, agent productivity, automation deflection, salary costs, software costs, escalation rates, and customer satisfaction impact. * Every assumption lives as a labeled, editable input on the Assumptions sheet — never hardcoded inside a downstream formula — so the COO can pressure-test staffing and automation decisions live. Sheets needed: * “Assumptions”: all editable inputs grouped by Customer Growth, Ticket Demand, Channel Mix, Agent Productivity, Automation, Cost Structure, and SLA Targets * “Ticket Forecast”: monthly ticket volume by channel, including new-customer onboarding tickets, recurring support tickets, escalations, and deflected tickets * “Capacity Plan”: required agent hours, available agent capacity, hiring needs, utilization %, backlog risk, and SLA coverage by month * “Cost Model”: support labor cost, software/tooling cost, outsourcing cost, automation investment, cost per ticket, and cost per customer * “SLA & CX”: first-response time, resolution time, backlog, CSAT estimate, escalation rate, and churn-risk indicator * “Dashboard”: single-screen executive summary with KPI tiles for total tickets, required headcount, support cost/customer, SLA attainment, automation savings, and CSAT trend; include charts for ticket volume by channel, capacity gap, and support cost over time * “Scenarios”: Base / Growth Surge / Automation-Heavy toggle that flexes customer growth, ticket volume per account, automation deflection, and agent productivity Formatting: USD currency with no decimals, percentages where relevant, frozen header rows, clean executive-friendly palette, conditional formatting for SLA misses and utilization above 85%, monthly columns grouped by quarter, totals row on every model sheet, and named ranges for major assumptions so formulas are readable. Before building, list your top 10 assumptions so I can sanity-check them, then execute.” 创建一个 Excel 电子表格：面向快速增长的 B2B SaaS 公司、覆盖 2026 年第三季度至 2027 年第二季度的客户支持容量与成本模型。目的：构建一个可供管理层使用的模型，用于决定何时增聘支持客服人员、哪些环节最适合引入自动化，以及随着客户数量增长，支持成本如何变化。业务背景： * 该公司向中小企业（SMB）和中端市场客户销售项目管理软件。 * 当前客户基数：4,800 个账户，月环比增长 8%。 * 支持渠道：电子邮件、在线聊天、电话和自助帮助中心。 * 工单量取决于客户数量、新客户引导活动、产品复杂度以及季节性因素。 * 您需要自行设定以下假设：每个账户的工单数、渠道组合、首次响应时间目标、处理时间、客服人员效率、自动化分流率、薪资成本、软件成本、升级率以及客户满意度影响。 * 每个假设都作为带有标签、可编辑的输入项存放在“假设”工作表中——绝不要在后续公式中硬编码——以便首席运营官可以实时压力测试人员配置和自动化决策。所需工作表： * “假设”：所有可编辑输入项，按客户增长、工单需求、渠道组合、客服人员效率、自动化、成本结构和 SLA 目标分组 * “工单预测”：按渠道划分的月度工单量，包括新客户引导工单、常规支持工单、升级工单和被分流工单 * “容量计划”：所需客服人员工时、可用客服人员工时、招聘需求、利用率%、积压风险以及按月份统计的 SLA 覆盖情况 * “成本模型”：支持劳动力成本、软件/工具成本、外包成本、自动化投资、每个工单成本以及每个客户成本 * “SLA 与客户体验”：首次响应时间、解决时间、积压量、客户满意度 (CSAT) 估算、升级率以及流失风险指标 * “仪表盘”：单页高管摘要，包含总工单数、所需人数、每位客户支持成本、SLA 达成率、自动化节省金额以及 CSAT 趋势等 KPI 卡片；并包含按渠道划分的工单量、容量缺口和支持成本随时间变化的图表 * “场景”：基准 / 增长爆发 / 自动化加强切换开关，可灵活调整客户增长、每个账户的工单量、自动化分流率和客服人员效率格式要求：货币格式为美元，无小数位；百分比格式；冻结标题行；采用简洁、对管理者友好的配色方案；对 SLA 未达标和利用率超过 85% 的情况设置条件格式；月度列按季度分组；每个模型工作表均包含汇总行；为主要假设定义名称范围，使公式更易读。在开始构建之前，列出您的 10 个最关键的假设，以便我进行合理性检验，然后再执行。

I just received a multi-tab Excel. 我刚收到一个多标签的Excel文件。

And the magic button is the Google Drive integration: 其中的神奇按钮就是 Google Drive 集成：

Click here to open your file inside Google (docx, slides, sheets…). Much better than having a file inside your computer.

It automatically created this Google Sheet. Perfect.

And if you need to edit it, you have two choices from here: 如果需要进行编辑，你可以从这里选择两种方式：

Go back to Cowork and ask what you want. And if you don’t know what to ask, ask Claude to ask you. Using the AskUserQuestion tool I told you about.返回Cowork并输入你想要的修改内容。如果不知道该怎么提要求，可以让Claude主动向你提问——用我之前提到的AskUserQuestion工具。
Or use the (free) ChatGPT extension inside Google Sheets. Here’s how: 或者使用Google Sheets内置的（免费版）ChatGPT扩展程序。具体操作如下：

Go to the top → Extensions → ChatGPT (you might have to add it first). 点击顶部菜单 → 扩展程序 → ChatGPT（若未安装需先添加）。

It will open a sidebar. I like using the Heavy mode. 它会打开一个侧边栏。我喜欢使用重度模式。

I tested every single ways to create spreadsheets with AI. Cowork is the best. 我测试了所有用AI创建电子表格的方法。Claude Cowork 是最好的。

And it’s not even the only cool part of Cowork. 而且这还不是 Cowork 唯一酷炫的地方。

You can onboard anyone of your company much more easily with Claude: 你可以通过 Claude 更轻松地让公司的任何人上手使用：

VII. Onboard with Claude.七、让Claude入职。

Go to your Claude app. Create a new project. 打开你的Claude应用，创建一个新项目。

Add all of the necessary resources, usually the company’s mission, deck, some interviews of the CEOs or C-levels… 添加所有必要的资源，通常包括公司使命、演示文稿、CEO或高管的部分访谈等。
If you can add best practices from support, they usually cover everything about the business a new employee should know. 如果能加入客服部门的最佳实践，这些通常涵盖了新员工应了解的关于业务的所有内容。
Invite them to the new Claude Project. 邀请他们加入新的Claude项目。
Create a skill on top of it that is called /answer-onboarding with how to use the data inside the Claude Project to answer anyone doing an onboarding.在此基础上创建一个名为/answer-onboarding的技能，说明如何利用Claude项目内的数据来回答任何进行入职引导的人。
Now as soon as a new recrue needs an answer, they start by going to the Claude Project, and do /answer-onboarding + ask the question.现在，一旦有新成员需要解答问题，他们首先进入Claude项目，输入/answer-onboarding并提问。
Claude will give a great answer. Claude会给出出色的回答。

I have a rule in my company: 我在公司定了一条规矩：

Please ask me as many questions as possible. Failing is winning. 请尽可能多向我提问。犯错即是进步。
Never ask me a question before giving me a potential answer. 在向我提问前，请先给出一个可能的答案。
Find the potential answer by doing what I just said: project + skills invoking how to (potentially) answer the new person.通过我刚刚所说的方法来寻找可能的答案：结合项目与技能，思考如何（可能）回答新人提出的问题。

It forces people to go from “I’m new, I can’t think for myself” to “I work actively to find the answer, but I still need help”.它迫使人们从“我是新手，我不会独立思考”转变为“我主动寻找答案，但仍需要帮助”。

It’s very different. 这有很大不同。

After this you’ll want to run everything through Cowork. 之后你会想把所有事情都交给Cowork处理。

Don’t.但别这么做。

VIII. When Cowork is the wrong tool.八、当协作工具（Cowork）选错时

You should use… 你应该使用……

Chat: nothing to build, nowhere to keep it, a question, a quick rewrite.聊天模式：无需构建内容、无存储需求，仅用于提问或快速改写
Claude Code: the deliverable is software, code, something to build.Claude代码模式：交付物为软件、代码或需构建的内容
Claude Cowork: everything document-shaped (decks, analysis, client work).克劳德协同工具：所有文档形式的内容（演示文稿、分析、客户工作）。

Claude is extremely powerful. 克劳德功能极其强大。

So powerful that its biggest blocker is most of the time… you. 强大到很多时候最大的障碍……是你自己。

IX. You are slowing us down.第九条。你拖慢了我们的进度。

This is a typical Cowork session: 这是一次典型的 Cowork 对话：

You type a prompt. 30 seconds. 你输入一个提示词。30 秒。
Cowork reads your skills. 50 seconds. It generates a plan. 50 seconds. It asks you clarifying questions using AskUserQuestion. 50 seconds. Cowork 读取你的技能。50 秒。它生成一个计划。50 秒。它使用 AskUserQuestion 向你提出需要澄清的问题。50 秒。
Now you answer those questions. You click some options. Fine. 现在你来回答这些问题。你点击一些选项。可以了。

But sometimes you need to type a custom answer (and custom answers are where the best outputs come from). 但有时你需要输入自定义答案（而最佳输出往往来自自定义答案）。

So you stop. Think. Type. 所以你停下来。思考。输入。

60 seconds per answer. Maybe 2 minutes. 每个答案花60秒。也许2分钟。

Across 8 questions, that’s 8-15 minutes of you being the slow part.8个问题下来，你就成了那个拖慢节奏的部分——耗时8到15分钟。

But Cowork can read 100,000 words in 15 seconds. It can build a spreadsheet in 90 seconds. And it has to wait for you to type at 60 words per minute. 但 Cowork 能够在15秒内阅读10万字，90秒内构建电子表格，却要等你以每分钟60字的速度打字输入。

But you could be faster. Because you speak at 150 words per minute. 但你可以更快——因为你的语速能达到每分钟150字。

Side note ⚠️温馨提示 ⚠️

I know it sounds ridiculous to optimize everything for speed. 我知道追求极致速度听起来有些荒谬。

But talking instead of typing has another massive benefit: it sounds natural. It’s you, your voice. 但用说的代替打字还有一个巨大的好处：听起来很自然。那是你，你的声音。

And your brain thinks very differently when it has to talk. 而且当大脑需要开口说话时，它的运作方式会截然不同。

Did you realize how good your ideas are, your flow is, when you’re talking to a colleague about solving a problem? 你有没有发现，当你和同事讨论解决方案时，你的想法有多精妙，思路有多流畅？

We want the same flow state here. 我们要的正是这种心流状态。

How to set up Wispr flow. 如何设置Wispr Flow。

Quick reminder for those who missed it: Wispr Flow is a dictation tool.温馨提醒错过的小伙伴：Wispr Flow是一款听写工具。

You hold a key, talk, release. Your words appear wherever your cursor is. 按住一个键，说话，松开。你的话语就会出现在光标所在位置。

Anywhere on your computer. Including inside Cowork’s chat box. 电脑上任何位置都可以。包括Cowork的聊天框内。

What makes Wispr Flow different: its accuracy. 使Wispr Flow与众不同的是：它的准确性。

Near perfection, every single time. 近乎完美，每一次都如此。

Wispr Flow + Claude Cowork is the perfect match to counter a common problem when using AI: you steer the conversation, and make sure you are in flow state. Wispr Flow与Claude Cowork结合，是应对使用AI时常见问题的最佳拍档：你掌握对话方向，确保自己进入心流状态。

Instead of typing “I need a Linkedin post,” you start talking “I recently found out about… and I want to share more about… but first I need to make sure that… so maybe we should start covering… and end up with… as a conclusion”.不再需要输入“我需要一篇LinkedIn帖子”，而是直接说出“我最近了解到……想要分享更多关于……但首先需要确保……所以我们不妨先探讨……最后以……作为总结”。

1. The initial prompt: I speak it. 1. 初始提示：我口述指令。

I say this out loud. Wispr types it. 我将它大声说出。Wispr 将其转化为文字。

The point isn’t to be faster, but we end up giving much more context when we talk (we are yappers by nature), rather than with a lazy typed prompt.重点不在于追求速度，而是当我们开口说话时（人类天生就是”话痨”），往往会比随手输入提示给出更多上下文信息。

And the more context, the better. 而上下文越丰富，效果就越好。

2. AskUserQuestion answers: I speak those too. 2. AskUserQuestion 的回复：这些语言我也说。

Cowork generates a form. Most options I just click. Cowork 生成一个表单。大部分选项我直接点击。

But when I need to add context (”make it more direct, she’s a CEO who hates fluff, and reference the ROI data from the last call”), I dictate that instead of typing it.但当我需要补充语境（比如“语气更直接一些，对方是位讨厌空话的CEO，并引用上次通话中的ROI数据”）时，我会直接用语音口述，而非手动输入。

I don’t self-edit while speaking. I dump my thinking. 我说话时不会自我编辑。我会一股脑儿地把想法倒出来。

Cowork figures out what matters. 协同工作能抓住重点。

3. Feedback and pivots: spoken. 3. 反馈与调整：口头表达。

When Cowork produces something that’s off, I used to type feedback like: “Tone is wrong. Make it less formal.”过去，当协同工作产出结果不符合预期时，我常打字反馈：“语气不对，别那么正式。”

Now I say: “The tone is too stiff. I want it to sound like I’m texting a friend who happens to run a 200-person company. Keep the data but make it casual. Only redo section 2.”现在我会说：“语气太生硬了。我希望听起来像是给一位恰好经营着200人公司的朋友发短信。保留数据，但口语化一些。只重写第二部分。”

Spoken feedback is richer because I am a yapper. 语音反馈更丰富，因为我话多。

How to download Wispr Flow for free. 如何免费下载Wispr Flow。

Go to wispr.ai. Download Wispr Flow. Install it.前往wispr.ai，下载Wispr Flow并安装。
The free tier is capped at 2,000 words/week - perfect to know if you like it. 免费版每周限2000词——足够让你体验是否喜欢。
Choose your favorite keystroke to activate Wispr Flow. I personally love “Shift”, the little arrow below “Enter”.选择你最喜欢的按键来启动Wispr Flow。我个人偏爱“Shift”键，也就是“回车”下方那个小箭头。
Go to any app, hold your keystroke (like shift) and… talk. 打开任意应用，按住你选的那个键（比如Shift），然后…开口说话。
You now type 4x faster than before. Congratulations. 现在你的打字速度比之前快了4倍。恭喜你。
You get a free month when you pay for my newsletter (you don’t have to).订阅我的新闻简报即可免费获得一个月使用权（当然，这不是强制要求）。

There’s no “integration” to configure. Wispr types wherever you type. 无需配置任何“集成”。Wispr 在你输入的任何地方都能直接输入。

Here’s how it looks: 效果如下：

嵌入式视频

5:16

My Cowork literally spawned 5 running agents. Mini-Claudes. 我的 Cowork 直接生成了 5 个运行中的智能体——迷你版克劳德。

Now there might be a problem. 现在可能有个问题。

Youll end up using Cowork so much you will spend a lot of money. 你可能因为高频使用Cowork而花费大量金钱。

So here is how to save money in Cowork: 那么，如何在Cowork中省钱呢：

X - How to save money in Cowork.X - Cowork省钱攻略

The $20 paid plan of Claude will give you credits (called tokens).Claude 的 20 美元付费套餐会提供积分（称为代币）。

But you will use them up very fast. 但这些代币很快就会用完。

At the scale of a company, saving tokens = saving thousands of dollars! 在公司规模下，节省代币 = 节省数千美元！

This section is about saving as many credits (= tokens) as possible. 这一部分讲的是如何尽可能多地节省积分（即代币）。

1 - Restart your conversation. Don’t send a follow-up. 1 - 重新开始对话。不要发送补充说明。

First, Claude doesn’t count messages. It counts tokens. Every time you send a message, Claude re-reads the entire conversation history. 首先，Claude 统计的不是信息条数，而是令牌数。每次发送信息时，Claude 都会重新读取整个对话记录。

Message 30 costs 31x more tokens than message 1. 第30条信息消耗的令牌数比第1条信息多31倍。

When Cowork gets something wrong, you want to type “No, I meant…” and send another message. Don’t.当 Cowork 出问题时，你可能会想输入“不，我的意思是……”然后发送另一条信息。但请不要这样做。

Every follow-up stacks on top of the full conversation history. Claude re-reads all of it every turn. At ~500 tokens per exchange, 20 messages burns 105K tokens. 30 messages burns 232K. 每一条后续消息都建立在完整的对话历史之上。Claude 在每次交互时都会重新阅读全部内容。按每条消息约500个token计算，20条消息会消耗105K个token，30条消息则消耗232K个token。

Instead: click “Restart the conversation from here” on a previous message (much higher ideally so you do save tokens).正确做法：点击之前某条消息下方的“从此处重新开始对话”按钮（最好选择较早的位置，这样能有效节省token）。

It’s at the bottom right of everyone of your prompt. 该按钮位于每条消息的右下角。

In Cowork, you can also edit a previous message. 在 Cowork 中，你还可以直接编辑先前发送的消息。

Click on Edit on your own prompt. It will delete whatever comes next. 在你的提示词上点击“编辑”，它将删除之后的内容。

And you can edit your prompt to start over (from there). 你可以通过编辑提示词（从那里）重新开始。

So when something goes wrong, edit your prompt with a better prompt or restart the conversation higher. 因此，当出现问题时，用更好的提示词编辑你的提示，或者从更早的地方重新开始对话。

It’s the biggest hack! 这是最实用的小技巧！

2 - Start a fresh session every 20 messages. 2 - 每20条消息开启一次新会话。

Long conversations are token furnaces. 冗长的对话会消耗大量token。

One developer tracked his usage and found 98.5% of tokens were spent re-reading history. Only 1.5% went to the actual output. 一位开发者跟踪自己的使用数据后发现，98.5%的token都用于重新读取历史记录，仅有1.5%用于实际输出内容。

When a Cowork session gets long: ask Claude to summarize everything, copy it, start a new session, paste the summary as your first message. 当Cowork会话变得冗长时：让克劳德总结所有内容，复制总结，开启新会话，将总结粘贴为首条消息。

There is no specific prompt, but I like this one: 没有特别的提示词，但我觉得这个不错：

“This Cowork session is getting too long. I need you to create the best summary possible as an .md file so I can start a new Claude Cowork session without losing too much context.” “这个 Cowork 会话变得太长了。我需要你创建一个最佳的总结，格式为 .md 文件，这样我就可以开始新的 Claude Cowork 会话，而不会丢失太多上下文。”

You keep the context. You lose the bloat. 你保留了上下文，去掉了冗余的部分。

3 - Batch your tasks into one message. 3 - 将你的任务批量整合成一条消息。

Three separate prompts = three full context reloads. 三个独立的提示词 = 三次完整的上下文重载。

One prompt with three tasks = one reload. 一个包含三项任务的提示词 = 一次重载。

Instead of: “Summarize this article” then “List the main points” then “Suggest a headline” write: “Summarize this article, list the main points, and suggest a headline.”与其写：“总结这篇文章” 然后 “列出要点” 再 “建议一个标题”，不如写：“总结这篇文章，列出要点，并建议一个标题。”

Using Wispr Flow helps. 使用Wispr Flow会有所帮助。

4 - Use Sonnet (not Opus) for quick tasks.4 - 使用Sonnet（而非Opus）处理快速任务。

Grammar checks, brainstorming, formatting, short answers. Sonet handles all of this at a fraction of the cost. And technically, Haiku is even cheaper. 语法检查、头脑风暴、格式调整、简短回答。Sonnet能以极低的成本处理所有这些工作。从技术层面来看，Haiku甚至更便宜。

Save Opus + Extended thinking for the work that actually needs it. Drafts and simple tasks on Sonnet-Haiku free up 30-70% of your budget for deep work. 将Opus+扩展思维模式留给真正需要它的任务。将草稿和简单任务放在Sonnet-Haiku上处理，能节省出30-70%的预算用于深度工作。

What about “Mythos” or “Fable-5” level of Claude? 那Claude的”Mythos”或”Fable-5”级别又该如何使用呢？

Fable-5, called a “Mythos-level,” is even better than Opus. Fable-5，被称为“神话级”模型，甚至比Opus还要出色。

It is obviously for your hardest tasks possible. 它明显是为您最艰巨的任务而设计的。

I’ll be honest, I was spamming Fable-5 requests whenever I could. 老实说，我逮到机会就疯狂使用Fable-5请求。

But you know it’s the most token-intensive model. 但您要知道，这是令牌消耗最密集的模型。

So there is no right or wrong answer. I am paying $100 a month for my Claude, and I have never run into any problems. We don’t use it for coding, so we are using fewer tokens than typical people. 所以没有绝对的对错。我每月花100美元订阅Claude，从未遇到过任何问题。因为不用于编程，所以我们使用的token数比普通人少。

5 - Delete your ABOUT ME file. 5 - 删除你的“关于我”文件。

I wrote extensively on building your “about-me” file.我曾详细写过如何构建你的“关于我”文件。

Keep it. But turn it into a skill: 保留它，但将其转化为一项技能：

It’s actually very simple. Just do this and upload it.

And I can now invoke it whenever I want. 现在我可以随时调用它。

6 - Spread your work across the day. 6 - 将工作分散到一天中完成。

This one is impossible to actually do in practice, but it works. 这个方法在实践中几乎不可能做到，但确实有效。

Claude uses a rolling 5-hour window. If you burn your entire limit in one morning session, most of your daily capacity goes unused. Claude 采用滚动5小时窗口机制。若您在某次上午时段耗尽全部额度，当日大部分可用容量实际未被利用。

Split into 2-3 sessions: morning, afternoon, evening. 建议分2-3个时段使用：上午、下午、晚间。

By the time you come back, your previous usage has rolled off. 待再次使用时，先前已消耗的额度会自动重置释放。

And avoid peak hours (5-11 AM Pacific on weekdays) when the same query costs more against your limit. 同时避开高峰时段（工作日上午5点至11点太平洋时间），在此期间相同查询会消耗更多额度。

Cool in theory, but I never do it ahah.理论上听起来不错，但我从来不会这么做哈哈。

If I want to work, I work. 如果我想工作，我就直接工作。

That’s why I pay for the $100/month paid plan. 这就是为什么我支付每月100美元的付费套餐。

So far, no problem (but I had a lot on the $20 plan). 到目前为止没什么问题（但我用20美元套餐时遇到的问题可不少）。

I think a couple of dollars per day is worth it to have a godlike intelligence by my side. Your call. 我认为每天花几美元，身边就有神级智能相伴，这很值得。这取决于你自己。

XI. Your first 20 minutes.十一、你的前20分钟。

0–3: empty your settings, turn everything off. 0–3分钟：清空设置，关闭所有功能。
3–8: build your first Skill (something you keep doing). 3–8分钟：创建你的第一个技能（一件你反复做的事情）。
8–13: create your first Project (upload PDFs, excels… from a client). 8–13：创建你的第一个项目（上传客户的PDF、Excel等文件）。
13–17: use your skill inside your project to see how it feels. 13–17：在你的项目中使用你的技能，感受一下效果。
17–20: download my Claude Skills library by subscribing for free. 17–20：订阅免费内容，下载我的Claude技能库。

I give the library for free in my welcome email. 我的欢迎邮件中会免费提供这个技能库。

It’s all for free for new subscribers. 这对新订阅用户完全免费。

Creating skills should be your priority once you set up Cowork. 建立技能应该成为你设置Cowork后的首要任务。

If you want (yet) another way to create skills, here’s a cool last example:如果你还想（再）用一种方式创建技能，这里有个很酷的最后一个示例：

I saw this tweet praising a lot of experts.我看到这条推文盛赞了很多专家。

Go to Chat (not Cowork) and turn on “Research”. I then copied and pasted the tweet. 进入聊天（非 Cowork）并开启“研究”功能。我随后复制粘贴了推文。

It generated a long research on the topic after a deep web search. 它在深度网络搜索后，针对该主题生成了长篇研究报告。

Claude decided to create 7 skills out of it. Claude 据此决定创建 7 个技能。

Sky is the limit. Create many skills, test them, improve them. 潜力无限。多创建技能，测试它们，改进它们。

PS: You don’t want your company to be left behind. 附注：你肯定不希望自己的公司被甩在后面。

Reading about AI is useful. 阅读人工智能的相关内容固然有益。

But it won’t change how your company works. 但这并不会改变贵公司的运作方式。

If you have 100+ people on your team, adopting AI is not a “send them a newsletter” problem. It is a systems problem.如果你的团队有100人以上，那么采用人工智能就不是一个“发封邮件通知大家”就能解决的问题。这是一个系统性问题。

Your people have different roles, clients, tools, workflows, permissions, data, risks, and habits. My newsletter, sent to 750,000 people, can’t do that. 您的团队有不同的角色、客户、工具、工作流程、权限、数据、风险和习惯。我的新闻通讯订阅人数为75万，无法做到这些。

That is how I help companies. 这就是我帮助企业的方式。

I run tailored AI workshops for enterprise teams. I am not alone, we are 20 people in New York City to support our clients. 我为企业团队提供定制化人工智能工作坊。我不是一个人，我们在纽约市有20人共同支持客户。

We map where AI can actually save time, where it should not be used, and how your people can adopt it without creating chaos. 我们会规划出人工智能真正能节省时间的领域、不应使用的场景，以及您的团队如何在不造成混乱的情况下采用这项技术。

If your company has 100+ people, message me. 如果你的公司有100人以上，请联系我。

My team & I will take care of you in priority. 我和我的团队将优先为你服务。

A message from the author, Ruben. 作者鲁本的一条留言。

This article exists because 89,000+ people decided AI is too important to leave aside. Not only that, but they shared it around them. They understand they are the sum of the 5 people around them. So better have them using AI. 这篇文章之所以存在，是因为有超过89,000人意识到AI太重要了，不容忽视。不仅如此，他们还在各自圈子里分享传播。他们明白自己是由身边的5个人共同塑造的——所以，最好让这些人也在使用AI。

If this helped you — or if it’ll help someone you know — forward it to them. That’s how this grew. Just readers like you sending it to people like them. 如果这个对你有帮助——或者对你认识的人有帮助——请转发给他们。这是我们成长的方式。就是像你这样的读者，把它分享给像他们一样的人。

And if you’re new here, follow me on X →@rubenhassid (also free!)如果你是新来的，请在X上关注我 →@rubenhassid（也是免费的！）