Claude 技能详解：如何创建可重用的 AI 工作流

欢迎回到《How IAI》。

Welcome back to How IAI.

我是Claire Vo，作为产品负责人和AI狂热者，我的目标是帮助你们更好地利用这些新工具。

I'm Claire Vo, product leader and AI obsessive here on a mission to help you build better with these new tools.

今天，我们将开启一系列《How I AI》迷你剧集的第一期。

Today, we have the first of many How I AI mini episodes.

本周的主题将是Claude Skills——Anthropic最新推出的特性，任何人都可以创建并加载特定技能和工具，让Claude Code、API或claude.ai随时调用。

This week is going to be all about Claude Skills, the newly released feature from Anthropic that lets anybody create and load up Claude Code, the API, or claude.ai with specific skills and tools it can call on at any time.

我将向你们展示如何创建技能、什么是技能，以及一些关于如何在工作流程中使用技能的思路。

I'm gonna give you a view into how to create skills, what skills are, and a couple ideas about how you can use skills in your workflows.

我们开始吧。

Let's get to it.

本期节目由ChatPRD赞助。

Today's episode is brought to you by ChatPRD.

我知道，许多观众收看《How I AI》是为了学习如何实际应用AI，让构建过程变得更轻松。

I know that many of you are tuning into How I AI to learn practical ways you can apply AI and make it easier to build.

这正是我构建 ChatPRD 的原因。

That's exactly why I built ChatPRD.

ChatPRD 是一个 AI 副驾驶，能帮助你撰写优秀的产研文档、自动化繁琐的协调工作，并从一位专家级 AI CPO 那里获得战略指导。

Chat PRD is an AI copilot that helps you write great product docs, automate tedious coordination work, and get strategic coaching from an expert AI CPO.

它已被广泛采用，从增长最快的 AI 初创公司到拥有数百名产品经理的大型企业都在使用。

And it's led by everyone, from the fastest growing AI start to large enterprises with hundreds of PMs.

无论你是想快速原型开发、指导新手产品经理，还是在大型组织中高效扩展，ChatPRD 都能帮你更快地做出更好的成果。

Whether you're trying to vibe code a prototype, teach a first time PM the ropes, or scale efficiently in a large organization, ChatPRD helps you do better work fast.

我们已与你喜爱的工具集成，包括 v0.dev、Google Drive、Slack、Linear、Confluence 等。

And we're integrated with the tools you love, v zero dot dev, Google Drive, Slack, Linear, Confluence, and more.

因此你无需改变工作流程，就能借助 AI 加速进展。

So you don't have to change your workflow to accelerate with AI.

前往 chatprd.ai/howiai 免费试用 ChatPRD。

Try ChatPRD free at chatprd.ai/howiai.

让我们重新让产品工作变得有趣起来。

And let's make product fun again.

今天，我要和大家聊聊Claude技能。

Today, I'm gonna be talking to you about Claude Skills.

什么是Claude技能？

What are Claude Skills?

如何创建它们？

How do you create them?

对于产品工程师和设计师来说，Claude技能在日常工作中有哪些不错的应用场景？

And what would be some good uses for folks, especially product engineers and designers out there to use Claude Skills in your day to day workflow.

那么，什么是Claude技能？

So what are Claude Skills?

Claude代理技能是一组特定的指令和上下文，无论你是使用Claude Code、网页版还是桌面应用，都可以调用它们来完成特定任务。

Well, Claude Agent Skills are a specific set of instructions and context that can be called on by Claude, whether you're using Claude Code, the web, or desktop app to do a specific set of tasks.

Claude技能解决了一个AI使用者面临的有趣问题：即创建可重复使用的流程，这些流程包含一组指令，你可以根据对话上下文按需调用。

And Claude Skills solves a really interesting problem for anybody using AI, which is reusable workflows with a set of instructions that you wanna call on demand depending on the context of your conversation.

很多人可能会问：为什么不能用Claude项目来做这件事，或者像OpenAI的自定义GPT或项目呢？

Now many of you are gonna say, why can't I use Claude Projects for this or, you know, OpenAI custom GPTs or projects?

这些项目主要依赖于你加载到项目中的上下文。

Well, those really stick with the context you've loaded in those projects.

一旦你设置了项目，与这些项目相关的对话总会调用该上下文和指令。

Once you set up a project, those chats associated with those projects always call on that context and instructions.

它并不真正动态，你也不能按需调用项目来执行一组指令。

It's not really dynamic, and you can't call on a project on demand to get it to follow a set of instructions.

而且，我注意到Claude项目、OpenAI项目和GPT通常提供的是通用上下文，可以支持多种任务，但并不是专门针对特定任务的指令。

Also, what I've seen is Claude Projects and OpenAI Projects and GPTs generally have general purpose context that can feed a variety of tasks but aren't really task specific instructions.

因此，Claude代理让你能够明确定义特定任务的指令、示例，甚至可以运行脚本，让你的通用聊天机器人真正代表你执行任务。

And so what Claude Agents gives you is the ability to really define task specific instructions, examples, and even scripts you can run that allow your general purpose chatbot to really do tasks on your behalf.

我喜欢Claude技能的一点是，它完全是用自然语言实现的。

And what's nice about Claude Skills that I like is that it really is just natural language.

我们已经看到许多围绕代理的发布，都是基于工作流构建的。

We've seen so many releases around agents that are really workflow built.

如果这样，就那样，调用这个工具。

If this, then that, call on this tool.

作为一般的AI构建者，我的偏好是，你知道吗？

And my preference as a general AI builder is, you know what?

如果这些模型在自然语言方面如此出色，我们就应该能够用自然语言来定义事物。

If these models are so great at natural language, we should be able to define things in natural language.

因此，Claude Skills本质上就是包含指令、元数据和链接文件的Markdown文件，让你可以按需调用某个任务或技能，提供一组特定的指令，从而真正完成该任务或技能。

And so what Claude Skills essentially are are markdown files with instructions and metadata and linked files that allow you to call on demand a task or skill, give a specific set of instructions, and then really get that task or skill done.

现在，我来分享一些关于Claude Skills整体结构的观察。

Now here are some of my observations on the general structure of Claude Skills.

第一，我认为它们是定义和发现你反复使用LLM执行的任务的一种非常好的方式。

One, I think they're a really nice way to define and discover tasks that you're doing over and over and over with an LLM.

所以，如果你发现自己总是以某种特定方式分析数据、以某种特定方式创建文档、执行工作流程或运行脚本。

So if you find yourself constantly analyzing data a specific way, creating a document a specific way, going through a workflow, or running a script.

你希望这些类似微指令的内容能够长期保持一致。

You want these sort of like micro instructions that stay consistent over time.

我知道你们很多人手中都有一个Google文档、Markdown文件或GitHub仓库，里面保存着所有这些提示词，然后不断复制粘贴。

And I know so many of you have a Google Doc or a markdown file or a GitHub repository where you've just kept all these prompts and you're copying and pasting them in.

Claude Skills 为长期填写和重复使用这些任务提供了一个结构化的框架。

Claude Skills really gives you a structured framework for filling out and reusing those those tasks over time.

我认为 Skills 另一个做得很好的地方是通过相对文件引用将额外的内容和上下文打包到技能中。

I would say the other thing that Skills do quite nicely is bundle additional content and context into a skill through relative file references.

因此，一个 Claude 技能可以引用其文件夹中的其他文件，这些被引用的文件可以是示例。

So a Claude skill can reference other files in its folder, and those reference files can be examples.

它们可以是模板。

They can be templates.

它们可以是额外的说明，这有助于 Claude 更好地管理上下文。

They can be additional instructions, and it helps Claude manage context a little bit better.

你总是会获得代理指令，而在必要时，你还会获得从代理链接的上下文文件。

So you're always gonna get the agent instructions, And when necessary, you will get the contextual files linked to from the agent.

因此，我认为 Claude Skills 的发现机制和上下文管理非常有用，正如它们在帮助文章中所描述的那样，说明了上下文窗口的大小、使用的 token 数量以及按需调用的机制。

So I think the discovery and context management of Claude Skills, as you see here in their help in their in their help article where they're describing the context window, how many tokens it uses, and when it's called on demand, is very, very useful.

最后一点是，Claude Skills 可以附带可执行的 Python 脚本。

I would say the last thing is that Claude Skills can come bundled with executable Python scripts.

这可能更适合技术背景的观众。

And so this is maybe for more of the technical audience out there.

但如果你需要验证你的技能——比如你的技能在执行某种分析、数据清洗或技术实现时，能够在Claude技能中引用Python文件，这实际上是一个非常有趣的扩展，能让你不必依赖智能体或大语言模型本身来为你定义和一致地运行这些Python代码。

But if you want validation of your skills, if your skills are running some sort of analysis, data cleaning, technical implementation, the ability to reference Python files within a Claude skill is actually a really interesting extension and keeps you from having to rely on the agent and the LLM itself to define that Python for you and run it in a consistent way.

尽管包括Claude在内的许多工具在生成代码时都具备代码执行能力。

And so while many of these tools, including Claude have sort of code execution capabilities when they're generating that code themselves.

但你可以看到，这些脚本或代码的定义存在很大的差异。

You can see that you get high variability in the, you know, definition of these scripts or the code.

而能够编写出一致使用的可执行脚本，这些脚本能真实反映你想要编写的代码，是非常非常有用的。

And the ability to actually write out executable scripts that would be consistently used that you feel like are represent of the code you wanna write is very, very useful.

那么，Claude技能到底是什么？

So so what Claude Skills are.

为了真正总结一下，因为我已经描述了它们的功能和结构，Claude技能其实就是一个文件夹。

And to actually really sum it up, because I've described what they do and kind of the structure, Claude files is is is a folder.

我觉得我们在一些文档中还不够明确地说明，到底什么是Claude技能。

Like, I I don't feel we have been explicit enough in some of the documentation to talk about exactly what is a Claude Skills.

我认为这些核心模型提供商，包括Anthropic，虽然我很喜欢你们，但这里给你们一些建议：这些功能明显是工程师设计的底层组件，却期望普通用户能轻松理解。

One of the things that I think that these core model providers can do, including Anthropic, love you, but here's some tips for you, is these are clearly primitives built by engineers expected to be grokked by everyday people.

我会坐在这里，替你们把这些内容解释清楚。

And I'm gonna sit here and translate them for you.

所以，Claude技能这个实际对象，也就是你创建的东西，就是一个文件夹。

So a Claude Skills, the actual object, the thing that you make is a folder.

这个文件夹里包含一个skills.md文件，还可以有其他相邻的文件。

That folder has a skills dot m d file in it, and then it can have additional files next to it.

归根结底，设置一个Claude技能的方法，说真的，就是要么把这个文件夹放在Claude Code可以引用的位置，要么把这个文件夹打包成zip文件，上传到claude.ai网站。

So at the end of the day, how you set up a Claude skill, no joke, is you either put it in a folder for Claude Code to reference, or you zip up this folder and you upload it to the claude.ai website.

所以我只是想说，我其实花了整整五分钟，试图搞清楚这个技能的核心资源到底是什么。

So I just wanna you know, I spent actually a good five minutes trying to figure out, like, what the hell is the asset of of the skill.

而真正的资源就是一个Markdown文件，加上一组其他文件和文件夹，这些要么在你的本地目录中供Claude Code使用，要么被打包上传到云端。

And, the actual asset is a markdown file, a set of other files and folders, either, used by Claude Code in your local directory or zipped up and uploaded to the cloud.

这就是它们实际的工作方式。

So that's how they actually work.

现在我想谈谈这些文件的结构，因为当我们开始创建Claude技能时，了解文件内部内容非常重要。

Now I wanna talk about what's in the structure of those files because I think it's really important as we start to create Claude Skills, you know what's in the file.

所以我来逐条讲解一下相关的文档。

So I'm just gonna walk through some of the documentation on this.

每个技能都必须包含一个skill。

So every skill has to have a skill.

他们甚至会全大写地写出来。

They, like, write it in all caps.

S k I l l . m d 文件。

S k I l l dot m d file.

这将是你的提示和一组指令。

That's gonna be your prompt, your set of instructions.

现在，通用指令集和开放式语言指令集与技能之间的区别在于，技能包含一些结构化内容，你在创建Claude技能时需要了解这些内容。

Now the difference between the a general set of instructions and open language set of instructions and a skill is actually has some structured content in it that you need to know about when you're creating a Claude Skills.

首先你需要了解的是元数据。

So the first thing you need to know about is the metadata.

这是你技能文件顶部的一段内容，必须以这种特定格式编写，用于指定技能的名称和描述其功能。

And this is just a piece at the top of your skill file that you put in in this particular format that gives you the name of the skill so you know what it's named in a description, what it does.

这将帮助Claude中的智能体了解何时调用该技能。

So this is gonna help the agent in Claude actually know when to call this skill.

文件中第二个部分是指令。

The second thing that you're gonna have in that file is instructions.

这时就需要你运用所有的提示工程技巧了。

So this is where all your prompt engineering skills need to come into play.

你需要将提示指令和自定义内容以Markdown格式写入Markdown文件的正文中，AI非常擅长生成这种格式的内容。

You would put your prompt instructions, custom things in the markdown, in markdown formatting, which AI is very good at generating in the body of the the markdown file.

因此，在文件顶部是元数据，下面是提示指令。

So at the top, you have your metadata, and then you have your prompt instructions.

此外，还有资源和其他代码，可以被该技能的主指令文件引用。

And then there is resources and other code that can be referenced from that skill dot m d main instruction.

实现方式是为这些文件提供相对路径链接。

And the way you're gonna do that is actually give the content a relative link to that file.

所以是点斜杠加文件名，或者点斜杠加文件夹名再加文件名。

So little dot slash file name or dot slash folder name with the file name.

如果你不知道如何为代码定义相对文件结构，好消息是。

If you don't know how to define relative file structure for code, Again, good news.

AI 非常擅长这个，我们在生成 AI 代码时会给你一些示例。

AI is very good at that, and we'll give you some examples when we generate AI code.

但你可以引用与代理的 Markdown 文件同级的文件夹，也可以引用其他文件。

But you can reference folders next to the agent's agent's markdown file, and you can also reference other files.

所以，你的 skill.md 将是主提示，你还可以引用其他文件，在这些文件中放置额外的上下文，这些内容会在需要时由技能调用。

So again, your skill dot m d is gonna be your master prompt, and then you can reference other files in which you can put additional context that will be called as necessary by the skill.

然后，正如我所说，你要么将这些内容放在 Claude Code 可以使用的位置，要么将其打包并创建一个 Web 应用中的技能。

And then, again, as I said, you're either gonna put this in a place where Claude Code can use it, or you're gonna zip this up and create a skill in in the web app.

所以我想创建我的第一个技能，而 Anthropic 在他们的发布文档中确实提到，Claude 内部有一个创建技能的技能。

So I wanted to create my first skill, and Anthropic in their launch documents actually said that there is a create skill skill inside Claude.

我其实没找到它，但我猜如果进入聊天，我可以要求 Claude.ai 为我创建一个技能。

So I couldn't really find it, but I just guessed that if I went into a chat, I could ask Claude dot a I to actually create a skill for me.

于是我进入这里，在Claude网页应用中输入：帮我创建技能。

And so I just went in here and wrote in in the Claude web app, help me create skills.

它确实提到，在这个上下文中，技能是帮助我更有效地完成任务的专用指令集。

And it does say that skills in this context are specialized instruction sets that help me perform tasks more effectively.

因此，我觉得Claude内部似乎加载了一个专门的元技能，可以帮助你构建技能。

So it seemed to me like it does have a specialized meta skill inside loaded in Claude that can help you build skills.

我将带你浏览整个对话过程，因为在尝试用Claude和Anthropic构建技能时，确实有一些有趣的体验。

And I'm just gonna walk you through the chat because there were some interesting experiences actually trying to build a skill in Claude and Anthropic again.

我爱你，但遇到了一些棘手的问题，我会向你展示我实际为Claude生成技能的方法。

I love you, but ran into some sharp edges, and I'm gonna show you the way I'm really generating skills in in, for Claude.

于是我让它创建技能，并且按照惯例，我让它创建一个PRD生成技能。

And so I asked it to create skills, and on brand, I asked it to create a PRD generation skill.

在观察这个过程时，有趣的是，Claude Code这个用于创建技能的技能非常元层级，它确实似乎遵循了一个相当结构化的工作流程。

And so what was interesting about watching this, Claude Code skill to create skills, very meta, generate is it did seem to go through a pretty structured workflow.

因此，它让我对这些Claude代理如何思考使用技能有了更清晰的认识。

So it gave me a sense of how these Claude agents actually think about using skills.

它读取了一个示例技能，我推测这是技能本身指令中的内容。

It read an example skill, which I'm presuming is in instructions of the skill itself.

它查看了详细的示例，然后大致理解了需要做什么，并创建了这个 skill.md 文件，顶部再次包含了元数据。

It looked at detailed examples, and then it kinda understood what it needed to do, and it created this skill dot m d with, again, the metadata at the top.

我实际上要切换到 Markdown 视图，这样你可以清楚地看到它应该是什么样子。

I'm actually gonna switch into markdown views so you can see exactly what it's supposed to look like.

顶部有元数据，下面是一组指令。

Has the metadata on top and then a set of instructions here.

我从 Claude 技能生成器中学到的一点是，这些指令真的非常详细。

Now something I learned from the Claude Skill Generator is man alive are these instructions quite detailed.

所以即使你最终没有使用 Claude 中的这个流程来创建你的技能——我会展示我 elsewhere 使用的更高效的方法——你依然可以看到它心目中优秀技能的框架。

So even if you don't end up using this flow in Claude to create your skills, and I'll show you what I did elsewhere, which is a little bit more efficient, you can see the outline of what it thinks a good skill is.

它包括了何时使用该技能、执行技能时需要做出的各种决策的决策树、创建文档时可能遵循的模板（实际上非常长）、最佳实践，以及所有这些内容。

So it includes when to use the skill, what is a decision tree on on different types of decisions it needs to make about executing the skill, what's the template that it might follow when creating a document, which is quite long, actually, writing, best practices, all those sorts of things.

然后它还提供了几个示例 PRD 格式，供该技能选择。

And then it gave a couple example PRD formats for the skill to choose from.

因此，你可以非常明确地定义技能的功能，也可以更通用一些，或者提供大量选项。

So you can be very prescriptive about what a skill does, or you can be more general purpose or give it a lot of options.

这是我注意到的关于技能的一点。

That's something that I noticed about skills.

最后，它还提供了一些可以询问用户的问题。

And then finally, it gave some questions to ask users.

这一点很有趣，如果没在技能生成器里看到，我自己的提示里是不会加入的，但让智能体主动向用户提问以获取更多澄清信息，确实很有价值。

So this is something interesting I wouldn't have put in my own prompt if I hadn't seen it in this skill generator, but putting questions that the agent can actually ask users to get more clarifying detail is very interesting.

输出格式，即对技能本身的输出结果做出非常明确的规定。

Output format, so being very prescriptive about the output of the skill itself.

然后它给出了一系列示例。

And then it gives a bunch of examples.

你可以在底部看到两样我觉得有趣但不确定是否被使用的东西，那就是关键词。

And you can see at the bottom two things that I thought were interesting that I'm unsure if they're used, but very curious is keywords.

我推测这些是触发技能本身的关键词，这一点我写提示时根本不会想到，但非常有用。

So I'm presuming these are keywords that invoke the skill itself, something again that I wouldn't have thought of when writing the prompt, but is very useful.

然后是额外资源，它实际上链接了将要生成的示例文档，作为额外的上下文。

And then additional resources, it's really linking to the example documents that's gonna create as that additional context.

这是一个关于如何撰写优秀的Claude技能的绝佳示例。

And so this is a really good example of how to write a good Claude Skills.

所以，即使你并不真正使用这个流程，也值得生成一个，以便更好地理解构建优秀Claude技能所需的内容。

So even if you don't actually use this flow, it's worth generating just one to give you a good idea of what it takes to build a great Claude skill.

而当它偏离轨道的地方是，它创建了核心技能。

And then what where it kinda went off the rails is it created the core skill.

它创建了一个许可证文件，而我其实并不需要这个。

It created a license file, which I didn't really want.

它创建了一个快速参考指南，但这个其实用不上。

It created a quick reference guide, which is not really used.

它生成了一堆示例文件。

It created a bunch of example files.

它检查了文件大小。

It checked the file sizes.

我觉得，使用这个工具——至少在我使用Claude AI这个工具的经验中——它做了大量不必要的工作，让我根本没法真正使用这个技能。

Like, I think the challenge with using this tool or at least my experience of using this tool in Claude AI is it did a lot of unnecessary work that kept me from actually using the skill.

所以它创建了，让我想想。

So it created probably let's see.

一、二、三、四、五、六、七、八，一共12个文件。

One, two, three, four, five, six, seven, eight, 12 files.

我其实只需要5个文件，它却创建了12个。

It created 12 files when I really only needed five.

更糟的是，当我尝试下载这些文件时，它失败了。

And then the crummy thing is when I actually tried to download the files, it it failed.

它失败了。

It failed.

我收到了一个错误。

I got an error.

所以我不得不一个一个下载这些文件，把它们放进文件夹，再打包成ZIP上传，才能让它们真正生效。

And so I would have had to download these files one by one and put them in a folder and upload them in a zip to get them to be to be effective files.

所以我认为，使用Claude.ai来生成技能这一流程很有趣，能让我了解系统是如何被提示以生成一个优秀的技能的，但我不认为这是创建技能本身最有效的方式。

So I think this flow using claude.ai to generate a skill was interesting to figure out how the system is prompted to generate a good skill, but I wouldn't say it was the most effective way to create a skill itself.

那么，有什么更好的方法来创建Claude技能呢？

So what was a better way to create Claude Skills?

好吧，我告诉你，我发现用Cursor创建Claude技能是完成这些工作最简单的方式。

Well, I'll tell you, I found that creating Claude Skills in Cursor was the easiest way to get this stuff done.

所以我做的是，在我的本地机器上创建了一个完全空的文件夹。

And so what I did is I created a completely empty folder in my local machine.

我打开了那个文件夹。

I opened up that folder.

它叫做Claude Skills。

It's called Claude Skills.

我把那个文件夹在Cursor中打开，然后开始一个对话，内容是：为我创建一个用于创建Claude技能的代理/技能。

I opened up that folder in Cursor, and I started a chat that said, create me a agent slash skill for creating Claude Skills.

这是相关文档。

Here are the docs.

所以，我没有依赖Anthropic发布的Claude技能本身，而是直接给Cursor提供了关于技能结构的文档链接，并让它创建一个用于生成技能的技能。

So instead of relying on the Claude Skill itself that Anthropic put out, I just literally gave Cursor the link to the documentation about what the structure of a skill looks like and asked it to create a skill to create skills.

我必须说，这种方式比在Claude网页应用中创建快得多。

I will say this was much faster than creating it on Claude on the web app.

整个过程大概只花了三分钟，而我觉得在网页应用中生成可能花了十分钟，而且我还得不到文件。

It took probably three minutes where I think the web app took probably ten minutes to generate, and I didn't get the files.

它创建了一个不错的待办清单，你可以看到，它再次遵循了创建优质技能的指导原则。

And it created this nice little to do, and you can see, again, it's following the instructions for creating good styles.

所以，创建主技能.md文件、示例技能、模板文件和验证脚本。

So create the main skill dot m d file, create example skills, create template files, and a validation script.

你可以看到，所有这些都被放在了这里的create-skill文件夹中。

And you can see all of this was put in a create skill folder over here.

这个技能再次遵循了YAML格式，顶部是元数据，主体是指令。

And the skill followed again that YAML formatting with the metadata at the top, the instructions in the body.

它比Claude的技能用户生成的内容要短一些，但仍然链接并引用了其他文件。

It was a little shorter than what Claude generated with its Skills user, and it did link and reference other files.

你可以看到，这些模板文件是以一种非常整洁的方式生成的。

And you can see those template files were generated in a really nice way.

Cursor 在为我创建技能时做了一件让我意外的事：它生成了一个技能验证脚本。

One of the interesting things that Cursor did when creating a skill for me that I was actually surprised about is it created a skill validation script.

因此，这再次展示了你如何在技能中使用 Python——在技能开发过程中，你可以让它运行一个 Python 脚本。

And so, again, this is just an example of how you can use Python within your skills is you can actually, as part of your skill development, ask it to run a Python script.

这个脚本有点有趣，它会检查 YAML 格式、文件结构、文件是否存在，并进行一些内容验证。

This one is a kind of funny one where it checks the YAML validation and the file formatting and if a file exists and does little content content validation.

这是一种关于 Python 脚本调用能力的有趣元应用，但它确实有效。

So this was an interesting meta use of the Python script calling ability, but it works.

这很有用。

It was useful.

仅仅几分钟后，我就有了这个漂亮的 create-skill 文件夹，Claude 可以直接使用。

And so this just a few just a few minutes, and now I have this lovely create skill folder that Claude can use.

那么，我们该如何实际使用它呢？

Now how do we actually use that?

所以我现在很高兴。

So I'm now happy.

我已经设置好了我的Claude技能。

I have my Claude Skills set up.

实际上，我为了使用这个技能，启动了Claude Code。

Well, what I really did to actually use the skill is I fired up Claude Code.

正如我所说，这些技能可以在Claude Code、Claude桌面版、网页应用以及API中使用，不过今天我们不会深入讨论API。

So as I said, these skills are available in Claude Code, Claude Desktop, the web app, as well as in the API, although we're not gonna go into the API today.

因此，我想既然这些文件在我的桌面本地，我完全可以使用Claude Code来调用这些技能。

And so I thought because these files are local on my desktop, that I could just use Claude Code to call these skills.

我真的很想看看，如果我在当前目录下启动Claude并调用这个技能，它会不会直接运行。

And I really wanted to see if I just fired up, typed Claude in this directory, and called the skill if it would just call it.

它确实运行了。

And it did.

所以你可以看到，我输入Claude Code的第一个提示是：使用create skill技能来创建一个技能。

So you can see the first prompt I put into Claude Code is use the create skill skill to create a skill.

这事儿你连说三遍试试看。

This is, you know, say that three times fast.

用于将更新日志条目转化为面向用户的通讯。

For turning changelog entries into a user facing newsletter.

所以这是我每周为 ChatPRD 都要做的事情。

So this is something that I do every week for ChatPRD.

我会把所有技术生成的更新日志条目整理成一份面向用户的通讯，发送给我们的所有订阅用户。

I take all of our technically generated changelog entries, and I write a user facing newsletter that goes out to all of our subscribers.

这是一项有非常特定格式、特定输入和特定输出的任务。

And it's something that follows a very specific format, has a very specific input, very specific output.

我觉得这非常适合做成一个 Claude 技能。

I thought it'd be great for a Claude Skills.

所以 Claude Code 立刻就识别出了这个技能，这非常棒，帮助我创建了一个将更新日志转化为面向用户通讯的技能。

So Claude Code picked up that skill right away, which is really nice and helps me create a skill for turning changelog entries into a user facing newsletter.

所以它正在遍历仓库或文件夹。

So it's going through the repository or the folder.

它正在查找技能所在的位置，并找到了我的创建技能的 Markdown 文件。

It's looking for where the skills exist, and it found my create skill markdown file.

它读取了该文件，并理解了创建技能的结构。

It read it, and it understands the structure of creating a skill.

然后它为我的变更日志转通讯技能创建了一个目录，以正确的格式写入内容，用那个 Python 脚本验证了技能，并给我提供了一个摘要。

And then it went ahead and created a directory for my changelog to newsletter skill, wrote the content in the right format, validated the skill with that Python script, and then gave me a summary.

这个技能比之前我生成的技能简单得多。

And this skill is much simpler than the skill I generated before.

正如你所见，它只是一个单独的文件。

As you can see, it's a single file.

它在顶部有元数据，然后是一大段指令。

It has the metadata at top, and then it has a long set of instructions.

所以再次强调，创建技能时你并不需要使用所有这些链接的文件和文件夹。

So again, you don't have to use all these linked files and folders when you're creating skills.

你只需写一个优秀的提示词，就能完成一切。

You can really put in a good prompt and have that be that.

但现在，我的Claude Skills文件夹或仓库中又多了一个由我的元创建技能技能生成的技能。

But now I have another skill in this Claude Skills folder or repository generated by my meta create skills skill.

这个技能生成后，我实际调用了它。

And then what I did once that skill was generated is I actually called it.

使用这个功能特别有趣的地方在于，你不需要使用什么魔法词。

And what was really fun about using this is you don't have to use like a magic a magic word.

你不需要说类似‘调用Claude技能来执行a和c’这样的话。

You don't have to say like, invoke the Claude skill to do a, and c.

你完全可以直接说：好的。

You can literally I said, okay.

这是最新的变更日志。

Here's the latest changelog.

它唯一需要的关键词就是‘变更日志’。

So the only keyword it had was changelog.

有了这个上下文，Claude代理就能理解我的变更日志，并说：我要把这个技术性变更日志转化为一份引人入胜的通讯。

And with that context, the Claude agent was able to take my changelog and say, I'm gonna take this technical changelog and turn it into an engaging newsletter.

所以我根本没说把它转成新闻通讯。

So I didn't say turn it into a newsletter.

我也没说要调用这个技能。

I didn't say use the skill.

我只是说了一句‘这是更新日志’，它就结合我文件夹里的可用技能推断出了我的需求，还撰写了2025年10月的产品更新公告。

I literally just said, here's a changelog, and it inferred what I wanted based on the skills available in my folder and wrote a update for our October 2025 product updates.

现在回看这件事，有意思的点在于，我看完之后会觉得，可能我的技能里用的表情符号太多了，之后我就会反复修改这个技能，确保它完全符合我手头任务要求的格式和框架。

Now looking at this, what's interesting is I would look at this and I would go, maybe my my skill is a little too emoji heavy, and I would revise the skill over and over and over again to make sure it follows the format and structure that I want for the the task at hand.

但说到底，我只用了三分钟就创建好这个技能，然后大概花了一分钟就生成了这份通讯。

But at the end of the day, this took me three minutes to create the skill and probably one minute to generate the newsletter.

现在我拥有了一个可以随时调用的技能了。

And now I have a skill that I can use all the time.

我还在构思其他可能创建的技能，比如我可以做一个技能，能把客户演示内容转化为发给试用潜在客户的跟进邮件。

And just thinking for through other skills I might generate, I could do something like create a skill to turn customer demos into follow-up emails for trial prospects.

Claude Code 接下来要做的，还是调用那个技能创建工具技能。

What Claude Code is gonna do now is it's, again, gonna invoke that skill creator skill.

它会在左侧创建一个新技能，你已经看到过关于跟进的演示了。

It's gonna create a new skill over here in the left hand, which you see already demo to follow-up.

它会把指令填好，随时可以使用。

It'll put the instructions in and ready to go.

所以对于任何想开始使用 Claude 技能的人，这是我推荐的流程。

So for anybody that's trying to get started with Claude Skills, this is the flow that I would recommend.

创建一个你可以在电脑上访问的文件夹。

Create a folder that you can go into on your computer.

把它当作你的 Claude 技能仓库。

Make it your Claude Skills repository.

如果你想更高级一点，可以把它部署到 GitHub 上。

If you wanna get fancy and deploy it to GitHub, you can.

也许我会把这个技能部署到 GitHub 上，分享给大家。

Maybe I'll deploy this one to GitHub and share it with you all.

创建一个用来生成技能的技能，或者我再次把这个分享给大家，你们可以直接克隆我的技能。

Create a skill to create a skill, or maybe I'll, again, share this with you all and you can clone my skill.

然后让 Claude Code 帮你创建技能。

And then just ask Claude Code to make your skills.

当你在 Claude Code 的这个文件夹中时，你可以调用这些技能来完成各种你认为重要的任务，并随着时间推移不断提升这些任务的质量。

And then whenever you're in Claude Code in this, you know, folder, what you can do is invoke those skills to do a variety of tasks that you think are really important and improve the quality of those tasks over time.

这就是在 Cursor 中创建 Claude 技能的简单指南。

So that's the really simple guide to creating Claude Skills in Cursor.

我们来看一下。

So let's see.

它正在思考。

It's thinking about it.

它正在处理第 74 步。

It's taking 74.

我们等一分钟，看看我们的演示后续技能会返回什么结果。

We'll wait one minute and see what comes back for our demo to follow-up skill.

好的。

Okay.

所以再来说一下，我觉得这些只需要三分钟。

So about three again, I think these just take three minutes.

三分钟后，它写出了这个跟进技能的演示，我们可以在上面看到，它能把客户演示笔记转换成个性化的跟进邮件。

Three minutes later, it wrote this demo to follow-up skill, which we can see up here that transfers customer demo notes into personalized follow-up emails.

它正在对这个技能运行Python验证脚本，我觉得这有点小题大做，但你知道的，你开心就好，Claude。

It's running the Python validation script on that skill, which, again, I think it's a little bit overkill, but, you know, you do you, Claude.

现在我的文件夹里又有了另一个技能，可以随时调用。

And now I have again another skill here in my folder that I can invoke at any time.

所以这就是如何在大约三分钟内，通过一点创意编辑，为Claude技能再创建一个新技能。

So that's how in about three minutes, the little creative editing, you can get another skill for Claude Skills.

现在我想向你们展示的最后一项内容是，如何把这些技能实际导入到网页界面或桌面应用中，我知道很多人都在使用这些。

Now the last thing I wanna show you is how you actually get those into the web UI or the desktop app, which I know a lot of you are using.

所以，正如我所说，这些是需要上传到Claude的压缩文件。

And so, again, as I said, these are zipped up files that you need to upload into Claude.

因此，你需要打开你的 Finder 或者桌面的文件浏览器。

So what you would do is go into your finder or into your file browser in your desktop.

你需要将这个文件压缩成一个zip文件，而我其实之前没做过这个，所以我们现在现场操作。

You would zip up this file as you need into a zip, and then I'm gonna I actually haven't done this before, so we're doing it live.

然后你会把这个技能上传到功能模块中。

You would upload that skill into the capabilities.

这不行，因为我的技能名称只能包含小写字母。

Well, this didn't work because my skill name can only contain lowercase letters.

所以我要再试一次。

So I'm gonna try that again.

看起来技能名称需要使用连字符分隔的小写字母。

It seems like the skill names need to be hyphenated little guys in here.

所以我们再试一次。

So we're gonna just do this one more time.

我要保存这个。

I'm gonna save this.

我要更新我的技能为写技能，我们再试一次。

I'm gonna update my skill to write skills, and let's try this again.

好的。

Okay.

我要把这个压缩一下。

So I'm gonna zip this.

我要尝试把这个演示上传到Claude界面中的跟进技能。

I'm gonna try to upload this demo to follow-up skill in the Claude UI.

太棒了。

Bravo.

我成功了。

I did it.

好的。

Okay.

我得更新一下我的东西。

So I gotta update my thing.

现在我可以把客户演示笔记转换成个性化的跟进邮件了。

And now I can transform customer demo notes into a personalized follow-up email.

我现在可以在聊天中使用这个技能了。

I can then use that skill now in to my chat.

所以你可以看到，我刚刚添加了这个技能。

And so you can see here that I just added the skill.

干得好，Anthropic。

Good job, Anthropic.

实现了这个流畅的无缝工作流，直接集成到聊天中。

Doing that nice little seamless workflow into the chat.

它会读取这个技能。

It's gonna read that skill.

它可能会问我一些笔记，然后我可以把我最新的ChatPRD演示直接粘贴到Claude中，让它替我生成一封跟进邮件。

It's probably gonna ask me for some notes, and then I can take my latest ChatPRD demo and drop it right into Claude for it to create a follow-up email on my behalf.

这就是我们关于Claude技能的第一期迷你内容。

So that is our first mini episode on Claude Skills.

我谈到了Claude技能是什么，它们与在应用中提供上下文的其他方式有何不同，如何尝试用Claude创建Claude技能（这并不是我最喜欢的），如何用Cursor创建Claude技能（这是我最喜欢的），以及如何通过Claude Code或Claude网页应用上传zip文件来调用这些技能。

I talked about what Claude Skills are, how they are differentiated against other ways you can provide context in your app, how to try to create Claude Skills using Claude, which is not my favorite, how to actually create Claude Skills using Cursor, which is my favorite, and then how to invoke those skills using Claude Code or the Claude web app by uploading them as a zip file.

如果你喜欢这样的内容，请告诉我们。

If you like content like this, let us know.

我们将深入探讨如何使用这些AI产品中最新的功能。

We're gonna do deep dives on how to use some of the newest capabilities in these AI products.

我会分享一些我日常使用AI的方法，并为你提供最佳的AI工作流程，以提升你的工作和生活。

I'll share some of how I AI in my day to day, and we will get you the best workflows in AI to improve your work life and your personal life.

非常感谢你观看《How I AI》，我们很快再聊。

Thank you so much for joining How I AI, and we'll talk to you soon.

非常感谢你的观看。

Thanks so much for watching.

如果你喜欢这个节目，请在YouTube上点赞并订阅，或者更好的是，留下你的评论分享你的想法。

If you enjoyed the show, please like and subscribe here on YouTube, or even better, leave us a comment with your thoughts.

你也可以在Apple Podcasts、Spotify或你最喜欢的播客应用上收听这个播客。

You can also find this podcast on Apple Podcasts, Spotify, or your favorite podcast app.

请考虑为我们留下评分和评论，这将帮助其他人找到这个节目。

Please consider leaving us a rating and review, which will help others find the show.

您可以在 howiaipod.com 查看我们所有的节目并了解更多关于本节目的信息。

You can see all our episodes and learn more about the show at howiaipod.com.

下次见。

See you next time.