# FULL Claude Cowork Tutorial For Beginners in 2026! (Zero to PRO)

## Метаданные

- **Канал:** AI Foundations
- **YouTube:** https://www.youtube.com/watch?v=JdQ_FHgP5ms
- **Дата:** 04.05.2026
- **Длительность:** 40:58
- **Просмотры:** 19,209
- **Источник:** https://ekstraktznaniy.ru/video/49622

## Описание

💎 Community + Claude Automation Course: https://www.skool.com/ai-foundations
✅ Free business overview resource: https://aifoundations.io/resources/the-business-overview-generator

This video offers a comprehensive guide to Claude Co-Work, covering its fundamental features and advanced applications. Learn how to leverage claude ai for daily tasks like research and brainstorming, and discover powerful workflow automation techniques. We will show you how to use various ai tools to automate business operations, offering practical automation tips and tricks to boost your productivity.

--------------------
⌚Chapters:
0:00 - Claude Cowork Full Guide
0:42 - Pricing & Download
1:49 - Claude Cowork Features
3:48 - Prompting in Claude Cowork
7:55 - Cowork Sub Agents & Research
10:35 - The Results of Cowork Prompts
11:27 - The First Cowork Feature to Learn
11:56 - Claude Automation Course
12:59 - Live Artifacts in Claude Cowork
17:46 - Connectors in Claude Cowork
19:39 - Connecting Gmail to Claud

## Транскрипт

### Claude Cowork Full Guide []

In this video, I'm going to walk you through everything you need to understand about Claude Coowwork. From the basics all the way up to the more advanced techniques like creating automations that run while you sleep and live artifacts that connect to your data. Now, I use Claude every single day. I'm talking researching, brainstorming, scripting YouTube videos, and running operations in my business, doing things like creating proposals after Zoom calls. All automated, happening behind the scenes. Once you understand how to use it, you don't really need anything else. Everything you need is already built in. You don't need to connect 30 or 40 different tools or watch 10 different YouTube videos in order to get some results. So, without further ado, let's get into the Claude Co-work full tutorial. Now, let's take a

### Pricing & Download [0:42]

look at pricing. Because you can't use Claude Co-work on the free version of Claude, you at least will need the pro version of Claude. As you can see, this line item here says includes Claude Co-work. But I'm going to be honest with you, it is well worth it because Claude isn't only the builder. It's not only the thing building out your automations, building out your folder hierarchies, your systems that can run your business, do your work for you. It's not only building those for you, it's also completing the things for you. It's both the car and the engine. It's super powerful. And the fact that we can get that for $17 per month or 20 if you're not build annually is just insane. I'm on the $200 a month plan and I have been for 10 months now. I don't regret a single month. Now, the other unique thing about co-work is it doesn't live on claude. com. the web at all. You need to download the Claude Co-work application or the Claude desktop app. And so, what you need to do is go to claude. com/d download if you don't have this downloaded already. And then you can select download for Mac or you can download for Windows, whatever device you are on. and then we can finally get into co-work and get started. Let's quickly walk you through

### Claude Cowork Features [1:49]

the UI so you can figure out everything. What you're going to notice right off the bat is claw does not work in chats when it's in co-work. It thinks in tasks which is going to come in handy knowing that in the future because co-work is very outcome based. It's not a thinking environment like clawed chat is. Clawed chat is where you go to brainstorm, think, strategize. Cloud co-work get things done. Next we have projects. And projects are these little containers of information that co-work can hold. So you can have custom instructions in there. You can have automations. And this is just a little version of co-work that you customize to complete a certain job the same way every time. Next, we have the scheduled tab. And this is just where your scheduled tasks live. So when you get to the point in this tutorial where we are creating automations, what you can do is you can put this stuff on a schedule. So, you can have it run weekdays at 9:00 a. m. daily at whatever time you want. You get to set the schedule and Claude can go and do things. I gave an example here. Check my Google calendar for today's meetings and summarize my unwrite emails. Highlight anything urgent. And so, that's something that could run daily at 10 a. m. You could have a report on your desk in the morning. Next up, we have live artifacts. And live artifacts are pretty insane. This allows you to create mini apps in Claw that are connected to your real data. So, if you wanted to create a replica of Google Calendar mixed with your notion database to-do list, you could do that and the data would be updating onetoone two-way sync live. You can set it up however you want. We're going to be building one of these in this video. Next, we have dispatch. This allows you to use co-work from your phone. So, you can access the power of co-work from anywhere because your phone, it says, is like a walkie-talkie that can communicate with Claude on your computer. Super cool. And then we have the customize tab which allows you to create uh skills which are automations and connect to tools. And so those are all the things that we're going to be going over today as well. I just wanted to run you through that really quick. It's nice knowing where everything is so that you can properly use the application here. Now the first

### Prompting in Claude Cowork [3:48]

and most important thing to understand about Claude Co-work is you need to talk to it different and you need to operate with Claude Co-work Claude chat because Claude chat like I said is more of a brainstorming tool. This is where you go to really start getting some things done. As you can see, it says, "Let's knock something off of your list. " So, the best way to start in Claude Co-work is to create a folder on your desktop and then begin having a conversation with it. So, I'm going to select work in project and then I'm just going to select choose a different folder. These are all of my folders I've been working in, but I'm just going to click new folder and call this clawed co-work playground. This is the best and easiest way to get started is just create a folder on your system, call it your playground, and then hit create. And now you can hit open. And I'm going to give Claude access to this. It's going to ask you if Claude is allowed to read, edit, and permanently delete files in a specific folder. So, you do have to be careful. I wouldn't just give it access to your entire computer, especially if you're sending off these big commands. But I'm going to always allow access in this Claude Co playground folder. That's not a problem. But now that I'm in the folder, I can do anything I want. I could have a normal conversation. I could say something like, "What are some interesting facts from the year 1800 or something like that and I could send it off and it's just going to operate like usual. " Now, the cool thing about Clawude Co-work is it has a progress bar on the side and so it's going to show you the progress of longer running tasks and it's also going to create instructions for you as you go. So, this is just a normal chat thread and it's giving me some of these uh bits and pieces of information here. Now, let me just show you one of the cool things that Claude can do when you're using co-work. You can actually do something called spinning up sub agents. So, you can create these little subversions of Claude that go and research different items for you. I can just say create me a plan for how we can research the year 1800 with multiple different sub aents and we can go and do all of this research at the same time and I want to learn everything there is about this year. So tell me first what the plan is for this folder structure and then let's develop a plan for what research these sub agents should be doing as well. And so what this is going to do is since it's working out of that folder, Claude can also create folders and files within the folder that you put it in. So we could do like a research project. I also did something kind of unique here. I said, let's think about how we can use sub aents so we can do multiple pieces of research at once. I'm going to send this off. And now Claude should develop a plan for me. I'm using Opus 4. 7 with the adaptive thinking. Look at this. It's already creating a folder structure. I'd put everything under a new folder in your workspace/cloud coowwork playground/year1800 research. Inside that it's creating an index for us. So now it's creating an overview. Politics and government, wars and military, science and technology, arts and literature and music. It's creating all of these things that I can study from about the year 1800. And this is just a good example to show you the power of Claude Co and kind of how it operates. Maybe this isn't something that you would even find useful. Don't think about that. Think about what I'm doing right now. And think about what Claude's doing for us. It's creating an entire folder hierarchy that's going to live on our computer. It says each subfolder gets a readme with a scope plus an individual MD research file. AMD is like a glorified text file. It's markdown. So you can read it. Claude can read it. It's easy. And then it gives me the sub agent research plan. I'd run eight research sub aents in parallel. That means eight versions of Claude going and doing research all on their own. So you're not just waiting on Claude chat to find you one thing. You can deploy eight versions of Claude or is that however many you need all at once. And look what it puts these agents up to. It's telling me what these agents are going to go research. Agent two, agent three, agent four, agent five, and so on. It says, "Want me to kick this off as described? " I'm just going to say yes. Let's spin up all of those agents and let's get all of the information I'm going to need about the year 1800. Beautiful. I can send it off. It says

### Cowork Sub Agents & Research [7:55]

"I'll launch all eight research agents in parallel. Each will create its own subfolder, do web research, and write markdown files in your workspace. After they finish, I'll write the master synthesis report. " So, here I am in our Claude Co-work Playground folder. And right now, we just have that one folder that it created, the year 1800 research. But when I click in here, as you can see, it's already began creating different folders. It's researching a bunch of different things in Claude over here on the right in parallel. And look at all the times it's searching the web. It's doing this together. So, it's researching right now politics and government and arts and literature. And it's creating all of the information it could find. So, it created this file, Capital Move and the White House. It's giving me all the information backed with actual sources because it's actually using web search and researching things like Wikipedia and a bunch of these validated sources which is really cool. So I could use this information if I was like a teacher or something to create a lesson plan or to do whatever I want. But the point is this is one prompt and Claude's out there doing things for me now. It's actually doing an agentic loop, a research loop and writing these files to my computer which I can use later. I mean look at that. I'd already created all eight folders just like that. And if I click into these, these agents are in here working in these folders right now, all simultaneously from one prompt. Now, even in Claude Co-work over here in the right hand side, you can see what's happening in the folder without actually going to it. So, you can see all of the files it's creating. You can see literally everything. You can even click into them and read them live. So, you can get this nice view. This is Markdown, by the way. It's just an easy way to make headers, paragraph text, bullet list, and so on. It's getting into the economy, trade, and industry in 1800. And it's kind of creating this nice web of information. And if you dare, click the writing file button or the searching web, and you'll be able to see all of the results it's pulling in, all of the sources it's pulling its information from, and you can watch these agents run in parallel if you can keep up. It's a ton of information that it's pulling in. And I like that it's searching the web for everything and not just trying to think of it on its own from its training data, but it's actually going searching and uh creating things for us from this data, which is amazing. And that prompt was very simple. You could tell it to list every single source that it pulled from during the process and then go back and validate it with another sub agent. You could have like 16 agents validating each other and running it through different research loops with the data once it finds it. And this is the co-work. You don't see this in chat. You don't see all of this getting done in parallel. It does one file at a time, which in this case would have taken a long time, maybe all day. And take a

### The Results of Cowork Prompts [10:35]

look at this. All 64 research files are in place. These are documents Claude has created for my computer. Think about using this in any way you want. This is just an example to show you the possibilities of the power of claude coowwork and what it can do. And this is just a random, you know, history thing, but think about this with competitors. pricing. Think about this with a content strategy, a research strategy. It created about 70 documents for us, all organized with a couple of prompts. And so that's what I really want to point out is it organizes this in its own way. Multiple agents working at once. We have a nice massive document. So this is the co-worker when you're prompting and kind of the way you need to start thinking folders, files, actions, not ideas, thoughts, and chats.

### The First Cowork Feature to Learn [11:27]

The first feature that we are going to cover today with Claude Co-work that I believe is one of the easiest levers to pull in order to see a big return is live artifacts. Because live artifacts make all of that text living and breathing. It puts a face on things. It can update live. It's essentially a mini app that you get to build and you get to design that's connected to your desktop so you can visualize the files on your computer in a way you've never been able to with Claude Co-work. I'm going to show you that now. But before I show you

### Claude Automation Course [11:56]

that, in our AI Foundations community, I just released a full cloud automation course, which is really just like a Claude automation mastery. And there are updating modules that I'm creating every day in here where I'm showing you how to automate your entire business or your work life using Claude. So this is a bunch of juicy content where I show you how to discover your time and your revenue leaks. We created a tool for you to do so in the community and I also show you how to set up your automated seuite and I give you multiple prompts for working with claude in order to start automating your tasks every single day. I will leave a link in the description to join the community which gives you access to this full updating Claude course where I'm adding modules multiple modules every single week. Plus, you get access to the wonderful community. There are some amazing people here and a calendar with Q& As's to any question that comes to your mind. Anytime you get stuck, we kind of have your back on that support. So, I'll leave a link to join that in the description if you are interested. But now, let's learn how we can use Claude artifacts. So, I want to actually stay

### Live Artifacts in Claude Cowork [12:59]

in this same Claude co-work chat that we were just in. The chat that created all 64 of those research files for us. And those research files are research on the year 1800. and everything that happened, the overview, the politics in the government, the wars in the military, the science and the technology. Let's say that I want to create an interactive environment where I can learn about specific years, and the year 1800 is just one of those years. I can create a mini application in order to do so. Let me create that real quick. I'm just going to say I want to create a live artifact of all of this data. I want this to be a dashboard and a nice color way that feels professional. I want the color way to be espresso brown and billiard green. I want the fonts to be serif headers with inter paragraph text. And I basically want this to be a fun interactive dashboard where I can come in and learn about certain years. But I just want this to be a beautiful learning environment where I can click around, change pages, change tabs, and have it update live. Maybe even have a section for flashc cards so I can learn and understand more about the year 1800. also create like a goal progress tracker in there so that I can track my progress on certain categories almost like it's my own study buddy for learning the year 1800 and I want this all to be in a live artifact and so the new way to code is just natural language as you can see I just typed out that big paragraph with my voice and this is the keyword I want you to pay attention to a live artifact I'm going to send this off and now that Claude has the context of this chat it's going to create something amazing for us. And remember, this does not have to be with generated knowledge. I just generated all of this knowledge out of thin air and told Claude to go research all of it. This could be with something that you already have like all of your social media posts. This could be with your analytics, your CRM, anything at all. And so, keep that in mind. But this is me creating a live artifact. Okay. So far, it's been working for about seven minutes creating this artifact. But remember, this is one prompt. And if you haven't seen this yet, I know you're going to be amazed when it happens because it amazes me every time, and I've seen this hundreds of times, these artifacts being created. It's insane what Claude can do. Okay. And it is finishing up. I'm going to select create. And now we have this dashboard that it just created for us with one prompt. So we have a study in 1800, a hinge between two centuries, romanticism rising, the steam patent expiring, a battery invented in an election that nearly broke a republic. It gives the nice stats like a world population, the largest city at the time, global life expectancy, US population, Britain cotton usage. Interesting. This is just the overview page. Look at all these tabs it created for me and a nice study session down here. So this is essentially like an app that we just built with markdown files. Has some nice crosscutting themes here. Very cool. Let's go look at the timeline. So it created a nice timeline with dates and what happened on certain dates of the timeline in the year 1800. A lot happened in this one year. Notable people as well. Uh we can look at notable people born. This is so sweet. It did all this on its own. A notable people born in 1800 died in 1800 alive at the time. Thomas Jefferson, Napoleon. That's like interesting to me that these people were alive at the same time together. This is just a little bit of what an artifact can do. Let me go to the flash card section. Okay, so we have a nice flash card system in here with my styling. And so we can even filter these flash cards, which is insane. Looks like there's 40 flash cards total. There's five flash cards in society and daily life in the year 1800. About how large was the world's population in 1800? We saw that on the homepage. about one billion people. Click to reveal. Roughly 950 to 1 billion. Beautiful. I can go to the next question. What was the largest city in the world? So, I basically created my own study system in here. And this is kind of cool because it's unique to me and what I want to study accomplish. Very nice. Then we can go look at my progress, which it created me a progress tracker because I asked for it. And so, I can select certain things in here that I've studied and that I've mastered for the year 1800. and that percentage updates live. You get the point. I just wanted to show you around this because this was a single prompt, no revisions. It created all of this for me. And now if I go into my live artifacts tab over here on the left, I can just open up the Chronicle dashboard, close the sidebar, and we now have a nice full page view of this and I can just work out of here. Now, the next

### Connectors in Claude Cowork [17:46]

most important thing to learn in Claude Co-work is the fact that you can connect to your tools, your live tools like Google Calendar, Gmail, the Microsoft ecosystem, HubSpot, Notion, Air Table, pretty much anything in the world you can think of, like things like Stripe as well. You can connect to that and access live data and take actions in those apps from your prompt bar. So, if I wanted to create a new contact in my CRM, I could just speak to Claude and it could update my HubSpot or my notion, wherever your CRM is. If I wanted to send an email, I could send it from Claude. I could read an email in Claude. I could draft the email I'm going to send in Claude. Claude is really like an operating system once you connect your live tools. So, let's get into doing that. Okay. So, if you open up your sidebar, what you're going to notice is that there's a customize tab in Claude Co-work. In order to connect your live data, you want to select that customize tab. And then underneath skills, you're going to see a connectors button. Connectors are your tools. So, I'm going to select that. And you're going to see that I already have some of my tools connected here. What I'm going to do is just disconnect from my Gmail so I can show you that. That's one of the more common connection tools. In order to see all of the options Claude has for tools to connect to, you can hit this plus button in your connectors panel and then select browse connectors. And here you're going to get a massive list of everything that Claude can connect to. It can connect to Spotify, Monday. com, Google Drive, Intercom, HubSpot, Microsoft, Canva, Miro, Notion, Zoom, Slack. I mean, come on, the list goes on. You can even connect to Zappier, which can connect to anything basically. You can connect to Uber Eats, Cash App, Superhum, Excaladraw. I mean, I'm just going to quickly scroll through here, and I'm sure your mind's already going to be spinning, and your eye might catch one of the tools that you like using, but anything you can think of, you can pretty much put in here. So, the easiest

### Connecting Gmail to Claude Cowork [19:39]

way to show you an example is with Gmail because we're all familiar with an email tool. So, in order to connect to a tool, you just have to hit connect. It's that simple. And it's going to walk you through a sign-in process based on the app. So, Notion's a little bit different than Gmail, which would be from Stripe. So, I'm going to hit connect. And what that's going to do is bring me to a screen where I can sign in to any one of my emails. I'll sign into one and then I will hit continue. I'll continue again. And just like that, you're connected. I had hit three buttons. And sometimes if it doesn't connect, you might just need to quit Claude and reopen it. If it said connected, chances are it did connect, but it might not appear in the interface that you are connected. So you can close claud. I'm going to go to the customize tab, connectors, and voila, Gmail is connected. So now when you're in connectors, you have a couple of different options that you get to choose from. So you have read tools and you have write tools. Read means it can read your data. Write means it can write your data. So it can create emails for you, delete things for you. Not in this case. It doesn't look like it can delete Gmails for you, but it can create a new label for you. list drafted emails and uh the read tools are just like searching email threads, listing labels and retrieving specific emails. So you get approvals and you can gate what Claude can actually do. You can entirely block Claude from taking certain actions with connectors. The options per tool are a little bit different, but you always get these three. Whether you want Claude to ask for your approval, block the action entirely, or just always

### Using a Connector in Claude Cowork [21:17]

allow it. And so now that I have Gmail connected, let me show you a live example of what it looks like using a connector. I'm going to go back. I'm still going to work out of my clawed co-work playground folder. And if you want to use a connector before you send off a prompt, what you can do is hit this plus button, hover over connectors, and you can see all of the tools that you're currently connected to. And so if they're on, you will be able to use them and take actions in them with the prompt bar. So I'm going to draft a prompt. I'm going to just say act without asking. And then I'll hit yes, continue. And then I'm going to say, "Hey, Claude, what I want you to do is use my Gmail tool that I have connected. I want you to search my inbox for all of my receipts that you can find from 2024 and 2025, and I want you to categorize them in folders on my computer in a way that makes the most sense. I want this to be set up for the future so I can eventually put this on a schedule and so that these receipts can be updated in a single file, a single source of truth per category. So, I kind of want this categorized out in any way that you think is best for organization and automation in the future from this Gmail inbox because this is where a lot of my receipts are going. I would also like to easily be able to collect running totals for certain categories. This is essentially just trying to help my bookkeeper out here. And so, that's the type of power you get when you connect a tool like Gmail, right? All of my receipts go to my Gmail. And so, I could send off a prompt like this. Claude could go actually use that tool, get all of the data, throw it back in a folder, and then we're going to be good to go. And I could even set up an automation, which I'm going to show you after this. But I'm going to send off this prompt here. And I'm just working in my Claude Co-work Playground folder. I do have all of that history data in there, but that'll be okay. I'm sure it will work itself out. I'm going to send that off. And the thing that's nice about Co-work is before it just dives in and completes a task for you, it always likes to ask clarifying questions, which is very good. It says, "What format should the per category files be in? " This is the single biggest decision for bookkeeper friendliness and running totals. It gives me a recommended uh a different one and a markdown file, which I've already done this before. That's why it says match existing skill. I'll just do the recommended route, the CSV file. How should the categories be split? Your existing skill uses 10 categories. I just want to say forget about my existing skill. I'm redoing this entirely. You can decide uh how everything gets best split. I really do want top level personal and business categorized out. And if you can't figure out whether one goes in personal or business, you can set it aside for me to figure out. Or you can spin up a sub agent and research the tool to decipher whether it is personal or business use. And so I can give suggestions. I don't even have to pick the answers it gave me for the question. But that's what I love about co-workers is it gives you options. Where should the new structure live? I could do a claude co-work playground finances folder. That'll work. and how aggressive should I be about what counts as a receipt. I'll do strict plus flag uncatategorized for review. But now it's going to search through my Gmail, find everything, take all this into account, and organize my receipts from the last 2 years. As you can see, it's using the Gmail integration, and it's actually doing search queries in my email that might signal certain terms that would be in an email, like thank you for your purchase or order or your receipt or things like order confirmation or invoice. It's searching all of these keywords in order to find it. And this is a very intelligent way of going about it. And this would take me days to do this alone. So, this is very cool. Now, I did mention this a little bit earlier, but when Co-work has long tasks, you can see the progress over here. And for this one, there's eight steps in total. If you're looking ahead and you don't like one of these steps, what you can do is you can actually comment or recommend a change about it. So maybe instead of doing this step, you want to comment request a change and say, "Nah, don't do this or do it in a different way. " And then you can cue the message up. All right. And so it categorized all of my receipts with the totals. We have the total right here. We have another business total, a personal total. We have the top vendors. So right here, it created that finances folder. So in my Claude Co-work Playground folder that we were operating out of, it created its own system. Here, I'm going to select finances master. Let's take a look at the Excel sheet it created us. Okay. Now, a lot of this might be blurred out since it is my actual stuff in here, but it created spending by category software subscriptions. And so, I could just hand this off to the bookkeeper. And this organized everything from 2024 and 2025, which is great. And so, that was just one example of using a connector, but you can connect pretty much any of your tools. That was just a read process as well. That wasn't even me drafting emails, but I could have read any emails I had from the past 24 hours and then drafted responses to them. So, I didn't really have to do that all on my own. I can have one place for everything to go.

### Creating Skills in Claude Cowork [26:09]

Now, the next feature in Claude Co-work that we are going to cover are Claude skills. Claude skills are automations that you can give to Claude to do something the exact same way every single time. Now, you don't only have to use this with generating knowledge, but you can also use this with your tools that I just showed you how to connect. So you can automate certain actions to happen in your Google Drive, your calendar, your emailing tool, your CRM, project management, whatever it may be. You can automate certain actions to take place so that it happens the exact same way every single time. Think about things like client onboarding, content repurposing, drafting newsletters, anything that you do the exact same way every time. You can set up a skill or an automation and claude and you can have it do the work for you. So that's what I want to show you next. Now, an example I always show people around a skill because it makes the most sense in people's mind is a daily briefing skill. Now, let's say I want to pull in all of my data from my Gmail, my Google calendar. If I have a task list, I could pull in my task list. And let's say when I come here in the morning, I want to type in slash daily brief. And I want everything to be laid out for me for what I need to get done on a specific day. Real events pulling in from my calendar. real to-do list from my notion, real calls that I have, real everything pulled from my actual tools. You learned how to connect your tools. You can go to connectors and connect anything you want. Now, let's think about how we can automate some of the actions within that. So, I'm going to tell Claude exactly what I want to create an automation around. If you want to create automations around your tools, like I said, make sure you connect those first. I'm going to go and I'm going to just speak out what I want. It's the best way to get things done. Now, hey, I want to create an automation and I want this to be a daily briefing automation. The goal of this would be to every morning when I come into Claude, I want to be able to type in slashaily-brief and then I want that to run the automation where it checks my Google calendar. It checks my Gmail and it catches me up on everything from the night before. What happened overnight? What do I have to do today? What emails respond to? I want it to catch me up on all of this. And I want there to be fallbacks for if nothing is there, then it just says like have a great day or something, but let's really think about how we can uh set up an automation that can run for me. Okay, it's that simple. I'm telling Claude what I want. I've already got my tools connected and it's going to ask me questions if it needs to. I'm going to send this off. And when you just say you want to create an automation, that's going to trigger include to use the skill creator skill and it will create that for you. So it's asking me what time window should overnight cover for emails. I'll just say last 12 hours. Which email should flag as needs response? Unread from real people. That'll work. Threads where I'm behind. Sure. How should the calendar section work? Let's just do today's events only. And what format do you want the briefing in? I'll just say structured sections. And it's that simple. My automation's getting created now. And so it created everything for me. The daily brief output, this is what it's going to look like whenever I run it. It says daily brief. It gives me the date overnight, what happened. It's going to check my Gmail for me automatically. Today's schedule, light schedule, needs response, worth knowing. It gives me everything that I need. So what I can do after it finishes is select save skill. When I save skill, it says in the upper right hand corner, skill saved. A skill is an automation. When I hit manage, this is now in my personal skills and it has a description and it has the entire workflow right in here for me. So, what I can do now is I can test it. I can go back and I can select new task and I can say slash daily and then I can tab over daily brief. Whenever I run this, it's going to check my calendar. It's going to check my emails and if there's nothing, it will tell me. If there's something, it will also tell me. So now I can go get my coffee. I can go do whatever I want. I always give the coffee example because I like making lattes. So I uh have a 10-minute process for that. So just getting caught up on all my stuff in the background is really nice. And now it's just running that automation that we just created. It's listing my Google Calendar events. It's searching my email thread. And it's doing all of this for me from this one command. And just like that, it got me everything that I needed. It says overnight quiet stretch. No real person email needing attention. Two packages landed at 216. A new keyboard, which yes, I am working on my new keyboard right now. It's sweet. And my Alaskan gold salmon. Beautiful. And so it got me my calendar, got me my worth knowing, 40 newsletters in marketing/shipping noise email skipped. Calm day get some rest since it knows it's nighttime. So, that's what the daily brief automation will do. And that's just an example of one automation. Imagine if you automate Zoom calls to your CRM, meeting notes, actions happening, follow-up reminders, follow-up messages, and emails. You can do everything like that from Claude.

### Claude Cowork Projects [31:18]

Now, the next thing that we are going to go over are clawed co-work projects. This takes folders, hierarchies, and organization, and even automations in Claude to an entirely new level. A clawed co-work project is basically like a little container of clawude that you customize to act a certain way every time you enter it. And it's a little bit different than a folder because you can specifically set up scheduled tasks that run in a project. And projects are really focused on instructions. How do you instruct Claude to act a certain way? So let's dive in. Let's create a project together. Now, in order to get to projects in Claude Co-work, open up your sidebar. Over here on the lefth hand side, there's going to be a projects tab. You can click on that projects tab and then this is where all of your little clawed containers are going to live. You can pin those if you want and they'll stay up top. You can edit the details. You can archive a

### Creating a Cowork Project [32:08]

project. But let's create a new one. In the upper right hand corner, I'm going to hit new project. You're going to get some options here. So, you can import a project. If you're already familiar with projects and you've used them in something like Claude Chat, you can actually import a project from Claude Chat, which is cool. You can use an existing folder. So, a folder on your desktop. You can set up a project that works out of a folder or you can just start from scratch where you set up a new folder. You give new files, new instructions. I'll select start from scratch. Then you can give your project a name. I'm going to call this blog generator and then go. Perfect. We're going to inject generative engine optimization into our blog posts from our YouTube videos. And then I'll tell Claude how I want to work in this project. I want to have this project be based around creating generative engine optimized blog posts around my YouTube videos that I can upload to Google in order to get search traffic from things like ChatGpt, Claude, Grock, Gemini. I really want when people are asking questions for my blog post to be surfaced. And so I'm going to be dropping in YouTube transcripts. And you're going to have to use my voice files and my business overview in order to draft these blog posts that push sales forward to the AI Foundations community. So, I'm telling it exactly how I want to generate these blog posts. These are instructions. I look at these as very minuscule things that you implement that point Claude in the right direction. So, I'm just telling Claude generally what is this project going to be for? the files and everything else in between the folder you're working out of. All of that is going to be more critical when it comes to forming Claude in the exact way you want it. So for files, there's a couple of files that I'm going to add here. And I've got a lot of files that are going to help that's going to help Claude write in my style, use my point of views in life in general, and really start sounding like me. So I'm going to select this plus button. And in my business folder, in my context, I have a business overview, which by the way, I have a free resource down in the description. If you want to generate your business overview, you can use the free resource on our website that we actually created for this thing in specific. What this does is it makes Claude interview you. I even have an install guide, and this is for complete free, by the way. It gives you seven sections where it will run you through and interview you. And I have a prompt here that you can paste into Claude. And Claude's going to help you create your business overview document which gives a project in co-work a ton of different instructions. And so that's the file that I just dropped in here. It's a complete overview of my business AI foundations. So it gets Claude up to speed on everything about my products, my ICP, anything that it needs. So I've uploaded that as context here so Claude can use that and it knows how to generate blogs to push my product sales forward. Next, I'm going to add another file from my AI Foundations folder, and it's actually in my marketing context, and this is my voice and my POV. So, my point of view is like hot takes that I have, any opinions that I have because Claude is very kind of just general, and it doesn't try to make anybody feel bad. good. It just kind of stays right in the middle. So, I want to inject my point of view so that my brand has a backbone in these blog posts and also my voice. I don't really need to drag in my life stories, but these are just some files that I keep on hand so that I can make my AI writing sound like me. I show how I generated those in our claw automations course in the community. If you're interested in generating those as well, highly recommend it. But now I can choose a project location. And so I can have this just run out of our Claude Co-work Playground folder. I will select create. And just like that, we created our project. And as you can see, it's very nice. We can chat with Claude just like normal over here. And this is using the co-work harness. So we still have that task based outcomes that we want the files being generated to our computer. Beautiful. We have instructions over here in the upper right hand corner. Scheduled tasks that we can set up right here. And then we have context. So the file that we're actually working out of on our computer. And another cool thing is Claude bakes in the memory right here. So, as you talk with it, it's going to start picking up things about you. Uh, so if you say like, "I like salmon. I don't like going to bed past 10 or anything like that," it will pick up those preferences and those little things about you. So, it knows not to steer you in those directions and it knows what you like, what you don't like, etc. But

### Using the Cowork Project [36:38]

let me just create my first blog post in here. And this could be like a spot where I just go to generate all my blogs. And so, let me add a YouTube video transcript because that would be the flow. YouTube video to blog. I'll select add files or photos. I'll select my YouTube folder and within here I have transcripts of all of my recent videos. I'll do my claw just released live artifacts video and update video that I created. And now I can say create a GEO styled blog post that is mentioning things about this update for my AI foundation's website. The ultimate goal at the very end for a CTA would be to join our community. But the main CTA is just to watch this YouTube video. I can give you that link and then I can put a colon here and then I will give it the link to that YouTube video from this transcript. I will just paste it right in here. And now I will send that off. And now we can start refining the project and we can just do a single task in here. This project is only for generating blogs. That's the purpose. Anytime I come here, all of the context, all of the files that get that gets created, all of my past blog post, all of that's just going to be completed in this one little container so I can keep everything nice and organized. Okay. And so it created the entire blog post for me. I can even select view my blog post and live review it over here. And so it gives me all my metadata. It gives me a title here, a description, the author, the date, the tags, everything I need. And then it gets me my nice clawed live artifacts header here. It was almost 11 pm when I got the dashboard working. I was supposed to be in bed. Instead, I was staring at a soft brutalist command center instead of clawed co-work watching a calendar event jump from Wednesday to Friday in real time as I dragged it around in Google calendar. This is the moment live artifacts clicked for me. And then it starts going into kind of what I did in the video. And this is actually something that happened in that video. And the cool

### Scheduled Tasks in Cowork Projects [38:30]

thing is since we are in a project here, if you wanted to, you could set up automations that schedule this out for you and you could have this stuff running all day long. It will just save the output to your computer. You can set up scheduled tasks. So, in the exact same way that I created that skill, the daily brief skill, you can name a task in here. So task name, you can give it a description for the project, and then you can give a prompt that you set to run on a specific day. And if you create a skill just like I showed you earlier for this project, you can type in slash and then you can hover over one of your skills. And when you select it, that means every day at 9:00 a. m. this skill is going to run. So you can create customized automations in a project that run on a schedule that import all of the information, all of the knowledge, all of the data that this automation collects from your tools, from the web, from research. You could have like an a research automation that gives you the AI news updates every hour if you want. All of it will be kept within the confines of this project so you know exactly where to go to get exactly what you want. So this is clawed projects. You can customize this in a million different ways. I honestly can't even begin to tell you what you could do with this, but the general idea is instructions. How do you want it to act? Point it to the right files. I said, "Use my voice files in my business overview. " So, what did I do? Well, I uploaded my business overview. I uploaded my voice, my POV. So, it's injecting all of that stuff into the blog post. Just some ideas here for context you can add. Okay. And that

### Final Thoughts [40:11]

completes the full Claude Co-work beginners guide for 2026. I hope you enjoyed and I hope you learned a lot in this video. If you did, please like and subscribe. I would highly appreciate it. I think like 95% of the people who watch this channel aren't subscribed. So, even if I could get an extra 5% of you, I would that would mean the world to me. So, hit that subscribe button if you want to learn with me, get on Q& A calls with me, my brother who's a co-founder and a lot smarter than I am with this stuff and hundreds of other people who are actively learning how to automate their business, work, and they're AI enthusiasts just keeping up with the tools, you know, just having fun chatting with each other, then join the AI Foundations community. That's where all the operators and heavy hitters are hanging out. But at the end of the day, I hope you learned a lot in this video. I'll see you in the next