00:00:00Every single day it feels like there's a new Cloud Code feature, a new tutorial, another thing
00:00:04that you're falling behind on. And this is extremely overwhelming if you're just getting
00:00:09started, especially if you don't come from a technical background. But today I'm going to help
00:00:13you fix that problem and show you what concepts I would start with if I was starting brand new,
00:00:19completely fresh today. We are going to cover 35 concepts across four different sections,
00:00:25starting with the things you have to know at the beginning and ending with super advanced
00:00:30power user type stuff that you don't need to know right now, but you should know that they at least
00:00:34exist. And by the end, you're going to have a clear roadmap that you can get started with right away
00:00:39and start making some real progress. And the first concept we're going to cover is what is the actual
00:00:45difference between the cloud.ai web app, you know, this chat GPT type interface in cloud code itself?
00:00:53Well, it's relatively simple. These two systems use the same brain. So I have Opus 4.6 running inside
00:01:00a cloud code. And right here in the chat menu, I have Opus 4.6. The difference is when I'm using
00:01:06cloud code, this brain has a body. It has hands. It can actually do things on my behalf, whether
00:01:11that's write code, whether that's access my emails, whether that's the interact with my computer itself.
00:01:16That's the biggest difference. Cloud code can actually do things for me by and large. Now,
00:01:22concept number two is about installing cloud code. And luckily for us, this is relatively simple.
00:01:26If you just Google cloud code install, it will bring you to the cloud code documentation.
00:01:31And all we have to do is copy and paste a single line of code into the terminal or if you're on
00:01:37Windows, something like PowerShell. Now, the one you're going to choose is depending on your
00:01:42operating system. So if you're using Mac OS, Linux or WSL, you will just copy this. And if you're
00:01:47using Windows PowerShell, you will copy this to find your terminal or to find PowerShell. You're
00:01:52just going to go to the search menu. Same thing in Mac. I'm on Windows type in PowerShell. It will
00:01:57give you a scary interface like this. You're just going to paste it, run this command, and then it
00:02:02will run you through the setup wizard. It's very simple. It's going to ask you to log in and you
00:02:06will use your actual subscription plan. And after you run through the setup to actually launch cloud
00:02:11code, you can just type in cloud. And you will see something like this. But now we move on to concept
00:02:17three, which is where do we actually use cloud code? Because there's actually a lot of options.
00:02:22Because while I can use cloud code in the terminal like this, I can also use the cloud desktop app
00:02:27as well. This is also cloud code. And then we have co-work here as well. What about co-work?
00:02:32And what about chat? Or what about something like VS code and IDE and integrated development
00:02:38environment where I do have the terminal up right here, same as this PowerShell situation,
00:02:44but I have the ability to look at my files and I have another window here, right? It's all very,
00:02:48very confusing. Again, especially if you aren't used to any of this. Well, the truth is it doesn't
00:02:53really matter all that much. At the end of the day, the most powerful version is the version that lets
00:02:58you work in the terminal, whether that's using something like VS code, which I will show you how
00:03:01to do today, or you're just using the terminal raw, like in PowerShell or something like that.
00:03:06But when you get started, if that's all too much for you and you're like, Hey, I just want the
00:03:11easy to understand interface inside the cloud code desktop app, or even co-work for most users,
00:03:1999% of the things they are going to do can be done in any of these spaces. You're not going to
00:03:23be losing out. That being said, I do implore you to try out the terminal, at least for the good
00:03:29beginning, give it a week or two. And if at the end of that time period, you're still like, ah,
00:03:34this is too much. It's just not for me, go ahead and move to these other ones, whether that's
00:03:38cloud code or even cowork, because you can always go from the terminal to these.
00:03:43It's a little harder if you get used to something that kind of has, you know, the bumper rails on it.
00:03:47And then you try to move to something more powerful. And as easy as it is for me to say, the terminal
00:03:53isn't as scary as it looks because at the end of the day, it's just a prompt window. We're just going
00:03:58to be prompting cloud code inside of the terminal in the same way that you would chat GPT on the web
00:04:03app. And like I said, we will use VS code for today because I think it's an easy stepping stone into
00:04:08the terminal environment. This is what is called an IDE, an integrated development environment.
00:04:13It's still the terminal, but it's just a more friendly way to view it because you can see over
00:04:18here I have this Explorer tab. So when we start working inside of a file, everything I create
00:04:22inside of the file, I can see here and I can actually open up over here and I'll be more obvious
00:04:27later in the video. Now using VS code visual studio code is totally free. If you search up VS code in
00:04:32Google and click the first link you will be brought here. The open source AI code editor. You're just
00:04:37going to download it, run through the install wizard. And that's kind of all you have to do
00:04:41from there. Just search for VS code, click on it, and you'll be brought to a page like this.
00:04:46You're going to go up to file, go to open folder, and then we're going to create a new folder today.
00:04:52This folder is where we're going to work. And just like you would work with any other project on your
00:04:56computer. So I'm going to go to new folder. I'm going to call mine 35 dash test. Click on it. It's
00:05:02select folder. And you'll see a page like this. Next, I want you to go up to the top where there's
00:05:08the three buttons hit terminal, go to new terminal. And now our terminal is open right here at the
00:05:13bottom, just like it was when we did PowerShell. And now to open Clod, I can just type Clod.
00:05:18And there we go. Clod code is up and running. So let's talk about concept number four,
00:05:25which is permissions. So right now you will notice it should be blank at the bottom. But if I hit
00:05:30shift tab, you will see some options pop up. What do all these things mean? Well, these are the
00:05:35different permissions. These are like almost think of it as security settings for what Clod can and
00:05:40cannot do to your computer, to the files on your computer without your permission. Now, when it is
00:05:46blank, we are on the default permission setting. That means it's going to ask your permission
00:05:50explicitly to edit files. If I want to accept edits on, it will automatically edit files without asking
00:05:56for permission. So it's a little bit quicker. But if it's going to use any bash commands, that means
00:06:00terminal commands, that means changes to your computer, think downloading a certain dependency
00:06:05or program and installing it, it's going to ask for your permission to do so. Then we have plan mode,
00:06:09which we'll talk about in a second. Now there's actually a third permission. But to do that,
00:06:14you need to open Clod in a specific way. So if we do slash exit, that gets us out of Clod code.
00:06:20And now if I do Clod and I do dash dash dangerously skip permissions,
00:06:26it's going to open up Clod normally, but you will see a new setting here called bypass permissions
00:06:33on. That means it can edit files, delete files, download things all without my permission. Kind of
00:06:38scary at first, but over time understand this is where most people end up sitting on because it's
00:06:44just quicker. I will say I've never had an issue with Clod code deleting any files that I didn't
00:06:49tell it to. Although, you know, there's always those theoretical stories in the aggregate though,
00:06:54this will save you so much time once you get comfortable. But if you're not comfortable with
00:06:57that yet, you don't want it to go crazy. Just keep it on accept edits on. Now, if you shift tab a
00:07:01couple of times, that will bring us to plan mode, which is the last concept we will cover in this core
00:07:06essentials section. So plan mode means that when I tell Clod code to do something like let's build a
00:07:11website, it's not just going to go out and do it on its own. It's actually going to think about it.
00:07:18It's going to come up with a series of steps we need to execute first. And what you will probably
00:07:23see here is just going to come back with some questions for us. It's essentially prompting us
00:07:28to get a better idea of what we want to build. So plan mode is the number one way for you to get
00:07:35better outputs from Clod code because it's going to make sure your prompt doesn't suck. The prompt
00:07:39you're going to come up with is going to have a number of holes in plan mode. Let's just kind of
00:07:44fill those in because Clod will ask us questions to do exactly that. So it's asking for our type
00:07:49of website. We'll say a landing page. It's asking for our stack. If I don't know, we'll just go with
00:07:56number one. And for our purpose, we'll say this is for a personal project and then we'll submit it.
00:08:04So just wanted to take a moment to talk about my Clod code masterclass because it is the number one
00:08:08way to go from zero to AI Dev, especially if you don't come from a technical background.
00:08:13Everything in this course is for the beginner and we focus on real use cases to help you learn how
00:08:20to master this amazing AI tool. I put out new updates every single week. And if you want to
00:08:25get it, just head to chase AI plus there is a link in the pinned comment. So now is the perfect time
00:08:30to talk about concept number six, which is your mindset when using Clod code in the planning page
00:08:35is a great time to discuss this. Now, right now it's asking me a few more questions about our
00:08:40website. And we're saying we want to make it about this fake online app called Argus, the social media
00:08:46intelligence app. And it's going to create this landing page for us. But when it comes to mindset,
00:08:51you need to start thinking of Clod code as a collaborator. And it's very easy to do so in
00:08:57the planning stage, because you have this back and forth, but where people get tripped up is in this
00:09:01back and forth. They're going to just do whatever Clod code says, and they're not going to ask it
00:09:06questions. And one example of us doing it the wrong way was when it asked us what
00:09:11tech stack did we want to use? We just said, Hey, we're going to use Next.js and Tailwind.
00:09:15Do you know what Next.js and Tailwind is? Probably not. If you've never done something like this
00:09:20before now, oftentimes Clod code will give you a recommendation and it's your best bet to go with it.
00:09:26But if you really want to learn Clod code and understand the fundamentals of building with AI,
00:09:31that is when you need to stop and tell Clod code to explain these concepts to you. And that is how
00:09:37you actually going to learn and not just be a caricature of a vibe coder, because you have to
00:09:42ask yourself if it's this easy to build things, what's the difference between you and the guy
00:09:48down the street who I can replace you with? Because he can ask those same questions too and just hit
00:09:51accept, accept, accept, build me a landing page. What's going to separate you from the pack and
00:09:55actually build your skills at a foundational level is asking Clod code these questions to explain
00:10:00things to you. It is the infinitely patient tutor and we need to treat it as such. So here's the plan
00:10:06Clod code came back with for our landing page. And this is what you can expect when you use plan mode.
00:10:10It's going to give you a pretty detailed outline of what it's planning to do before it actually
00:10:15executes. And you'll see some options here. You'll see yes, and bypass permissions, which means it's
00:10:20going to run what it planned and it's not going to ask for permission. Yes, manually approve edits.
00:10:25So it will begin creating things, but it's going to ask for your permission every time it wants like
00:10:29create or edit a file, which is very annoying. And then we have this one, no refine with ultra plan.
00:10:34Now ultra plan will be in section four when we go into some of the power user stuff, some of these
00:10:39like high level features inside of Clod code. But for now, we're just going to say yes,
00:10:43bypass permissions. And just like that, it's going to start building our website for us.
00:10:46And what you see over here on the left are all the files it's building. So this is what's nice
00:10:50about VS code, especially when you first get started. All the stuff that's being built is
00:10:55right here. So it's the same as if I actually went into my file explorer, opened up 35 dash test and
00:11:01looked what was inside. I'd see the same things. Now, one of the files Clod code created was the
00:11:05clod.md file you see over here on the left. And this is what we're going to talk about for concept
00:11:10number seven. So what is the clod.md file? Well, this is going to be inside of every single Clod
00:11:16code created project and think of it as the instruction manual, so to speak for Clod code
00:11:21regarding that particular project. This is something that's telling Clod code certain
00:11:26conventions and certain rules. You want it to always follow all the time when working inside
00:11:32this particular folder. Now Clod code will automatically create this when it's building
00:11:36your project. So you don't have to touch this too much and do understand there are a lot of techniques
00:11:41regarding clod.md. Some people like me will say less is more. Other people say put as much
00:11:46information here as possible that you think is relevant and are kind of like outside the box
00:11:50rules. All you need to understand at the beginning is that this exists and whatever you put in here,
00:11:56Clod code is going to reference every single time you tell it to do anything. So it's a very powerful
00:12:02tool, but you want to make sure what you do put in here, if you do edit it actually is important and
00:12:07it applies to virtually every single prompt. So Clod code created our web page. It said a dev server
00:12:12is started here on localhost. Localhost is like an internal web page. If you click on this, it will
00:12:18take you to the web page and I'm inside of Chrome right now, but it's not connected to the internet.
00:12:22This is all local. So this is what it built for us. Extremely ugly. And throughout this video,
00:12:27I'll give you different tips and techniques to actually making this better, but just understand
00:12:31this is where it's living right now. And we can actually interact with it just like we would a
00:12:34normal webpage. But before we get into that, let's talk about one of the most important concepts of
00:12:40Clod code and really large language models in general. And that is concept number eight,
00:12:44the context window. So the context window can be seen if we do forward slash context. And what we
00:12:51see right here is how much of the context window we have used. In this case, we've used 48,000 tokens
00:12:57out of 1 million tokens. And you're probably like, "Chase, what the heck does that mean? I don't even
00:13:01understand what tokens are, let alone what that really means. 50,000 of a million. Okay, what are
00:13:06you talking about?" So tokens are the currency of large language models and Clod code. Every single
00:13:12word is a token. So every single word here that we've given Clod code inside of prompt and every
00:13:17word it spit back out at us equals one token. One word, one token. A little bit different in reality,
00:13:23but just let that be your mental model. All the tool calls it uses cost tokens. All the code it
00:13:28generated cost tokens. Everything it does costs tokens. And the context window is our budget.
00:13:34For Clod code, that is 1 million tokens. It differs for each and every AI system. Now, why do we need
00:13:42to care about this? Well, first of all, if we fill it up, our session kind of ends and you've probably
00:13:47run into this with chatbots throughout the years where all of a sudden it says, "Hey, I don't have
00:13:50enough room." The session is ending. That occurs when you fill up the entire context window. But
00:13:55what's important isn't just filling up the context window. It's the idea of concept number nine,
00:14:01which is context rot. And that's the idea. As I fill up the context window with more conversations,
00:14:07more tasks for Clod code, the actual effectiveness of Clod code goes down. So the more I use it,
00:14:15the worse it gets, which means we want to keep an eye of where we're at on our context window.
00:14:22Especially since we can reset it at any time. I can pretty much start back at zero whenever I want.
00:14:27Now, this comes at a cost, right? When I'm using a normal chatbot, say I'm inside clod.ai and I
00:14:33start a new chat, that's kind of painful because everything we've been talking about, I want it to
00:14:38remember. I want it to remember my conversation. I don't want to start all over again. That sucks.
00:14:42But if I start all over again, Clod code works better. So what am I supposed to do? Well,
00:14:47luckily starting over inside of Clod code isn't like starting over inside of a chatbot. It's
00:14:51actually much better because Clod code, even if I start all over and I can start over by simply
00:14:57doing slash clear, my session is now reset. Clod code is living inside this folder. It's on my
00:15:03computer. It's on a chatbot in the cloud. Whenever I talk to it again and ask it questions about my
00:15:08website, it doesn't matter that we started a new chat for all intents and purposes because it can
00:15:13just look inside the code base. It can look at all these files. It has clod.md. It has the ability,
00:15:18almost like a human being, to go inside the folder and see what is what. And therefore, because of
00:15:25that, I always want to be resetting my context window if I can. Because why do I want to be
00:15:30working with a worst Clod code at 500,000 tokens for the sake of keeping a conversation going
00:15:35that I can just have it remember anyways, right? So there's few downside to constantly resetting.
00:15:42On top of that, as you fill up the context window, your prompts are essentially more expensive. You
00:15:48are using more of your usage because every single prompt is essentially sending everything that came
00:15:53before it. Now there's a caching system, which kind of alleviates some of that problems. But the point
00:15:58remains, if I'm down here at token number 800,000, that's really costing me more per prompt than if I'm
00:16:04at token number 50K or 100K. And you see a lot of people complain about Clod code usage. And this
00:16:10isn't the only reason they complain about it. But in large part, it's because a lot of people don't
00:16:13know how to manage their context window properly. So they're using more usage and Clod code isn't
00:16:18working very well. So you can see the importance of A, always having your eye on the context and B,
00:16:25always resetting it whenever it makes sense. As a rule of thumb, you don't really want to go past
00:16:30200,000 tokens if you can help it. There are going to be instances where, you know, hey, I just need
00:16:36to have a longer window for whatever reason. But if you're hitting 20%, you need to ask yourself,
00:16:42do I need to stay in the session? If you don't have a very good reason, you need to reset it. And by
00:16:47reset, all you have to do is what you saw here, which is forward slash clear. And if there is
00:16:52something in the conversation you want to bring forward, just ask Clod code to have a quick write
00:16:56up about what you were talking about, and you can copy paste it in the next session. Now, what you
00:17:00can see here for me, and this is going to be concept number nine, is my status line. So right here,
00:17:06you can see it says 35 dash test. You can see the actual model I'm using, and I can see my context
00:17:11window at all times, which says 2% here, and yours probably doesn't say that right now. But you can
00:17:16create that yourself so you don't have to constantly do forward slash context. You can just look right
00:17:20down here and see where you're at, which keeps you kind of on top of it. So to do that, you're just
00:17:26going to do forward slash status line. And you're going to write a prompt that says something like,
00:17:31hey, I want you to create a status line for me. That's totally persistent. That includes,
00:17:37you know, the folder I'm in the model I'm using in my current context window usage. Once you do that,
00:17:44it will create that for you, and you're just going to have to reset cloud code, and you'll have a
00:17:48status line of your own. So we just had a ton of concepts right there. Context, context window,
00:17:53the ability to clear, context rot, and the status line. Take those all together, and frankly, you
00:17:59will be well ahead of people who've been using cloud code for a while because people do not adhere
00:18:03to this principle nearly as dogmatically as they should. Now let's talk about some other slash
00:18:08commands while we're here, and this is our next concept, and that is rewind. So let's say you did
00:18:13do slash clear. You've been really on top of making sure that context window is as clean as possible,
00:18:17but you realized, oops, I really shouldn't have done that. I had something else I wanted to do.
00:18:22Well, if I do forward slash rewind, I can actually go back to previous sessions inside of cloud code.
00:18:30So I could go back to before I did forward slash clear and essentially bring that conversation
00:18:35back to where I was or the conversation where I said, hi. This will also include code changes.
00:18:41So if we were talking to cloud code and we made some changes to our actual website,
00:18:45and let's say I didn't like them, I can do forward slash rewind. It's an easy way to essentially go
00:18:49back to where you were because it has these almost auto save points. Concept number 14 is another one
00:18:54you're going to want to know, and that is forward slash model. So forward slash model lets me switch
00:18:59between all the different cloud code models. So we have Opus 4.6, Sonnet 4.6, Sonnet with a 1 million
00:19:06context as well as Haiku. So what you're going to want to do here kind of just depends on what plan
00:19:10you're on. If you're on the pro plan, 20 bucks a month, you're probably going to want to stick
00:19:14with Sonnet all the time. If you're on max 5x, you can get away with Opus a lot. Depends how much
00:19:20you're actually using it. If you're on 20x, you can go crazy with Opus all the time. The model is just
00:19:25going to depend on your users. That's just kind of what it comes down to. How much money are you paying
00:19:29in enthropic? When it comes to Haiku, you probably don't need to be using this at all. It is an
00:19:34extremely cheap, damn near free model. And its use cases are very specific and niche. So unless you
00:19:41know what you're doing, you don't need to play around with Haiku. And you'll also notice some
00:19:44other model related commands like effort. This is how much thinking cloud code was doing. It's going
00:19:48to default to auto. And again, this is kind of usage based because the higher the thinking level,
00:19:53the more tokens, the more usage. So just depends on your task. Now, the last core concept I want
00:19:58to talk to you about is Git. So we talked about rewind, right? The ability to have these sort of
00:20:05built-in save points with cloud code. Well, there is another sort of save point and that is Git. So
00:20:11Git essentially is another type of save point, but it's going to be on your computer. And it's pretty
00:20:16much saving your exact spot in terms of the code you've built. So I highly suggest you use this
00:20:22whenever possible. Cloud code is very spun up and intelligent when it comes to Git. So you can just
00:20:28say something like let's Git commit this. And if I say Git commit, it's saving it. This will come in
00:20:36handy later because when we start using things like Git hub, well, we want to take whatever save point
00:20:41we have with our code and push it to the cloud on GitHub. And to do that, you need to have committed
00:20:48the code. So think of Git commit as a save point. Rewind is great, right? Rewind is great in the
00:20:54context of cloud code, but as we become more mature and as we like to think about, hey, how am I
00:20:59eventually going to get this code out into the world? We need to start using Git. And that is a
00:21:04great stepping stone. And it's almost a secondary save point next to rewind. So you can feel a little
00:21:09bit more confident that your code is not just going to vanish or disappear. So now we're in section
00:21:15three, which is the toolkit. This is where we're going to start bringing in some outside tools,
00:21:19things like CLIs and MCPs, as well as introduce skills, which are one of the most powerful
00:21:25features of cloud code. And it couldn't happen at a better time because our website is ugly. I mean,
00:21:31just look at this thing. This is the definition of AI slop, boring, flat, purple gradients. How
00:21:37can we fix this? I mean, obviously we could have given it a better prompt, but there are some tools
00:21:42that can help us specifically with something like front end design. So the next three concepts are
00:21:46going to be skills, the skill marketplace and the skill creator skill. So skills are just text
00:21:52prompts. They're just a way to tell cloud code to do a specific thing in a specific manner. What
00:21:57we're looking at right here is the official anthropic front end design skill. And like you see,
00:22:02it's just text, no fancy code, nothing crazy going on, just a text prompt. So the front end design
00:22:10skill is the exact same thing as me taking this whole thing, copying it, pacing it into cloud code
00:22:15and saying, Hey, create a better front end design than what you just did and use these sort of
00:22:20guidelines, but I don't want to copy and paste it every single time. So instead I would just use
00:22:25something like the front end design skill by either doing forward slash front end design, or just saying
00:22:31in plain natural language, use the front end design skill. And then I give it my prompt. So let me show
00:22:36you how to do that. Now, the first thing you need to do is install the front end design skill. Now
00:22:40there's a number of ways you can do this. The first way is just to do slash plugin slash plugin will
00:22:46bring up the plugin design marketplace. And I can go to discover plugins, which you see right here.
00:22:52And I can just search for front end design skill. Once you see that you just select it and then you
00:22:57install it and you can confirm you installed it by just tabbing over to the installed section.
00:23:02And you can see right here, front end design plugin. Now, once you install the plugin, it will give you
00:23:07some sort of message saying to do something like reload plugins. You just run that and it will get
00:23:12the plugin installed. But at that point, the skill is installed in Claude code. Claude code knows the
00:23:19skill exists. You just have to invoke it. Like I said, a couple of different ways to do that. You can
00:23:24always invoke it by using forward slash whatever that skill is called in this case, front end design.
00:23:30I can say something like use the front end design skill and it will use it. Or it's smart enough.
00:23:38If I say something like, Hey, we're working on the front end for this webpage, let's do X, Y, and Z.
00:23:43It's smart enough to pick up on that sentence and say, Oh, we're working on front end design. Oh,
00:23:46I have a skill for that. Let me go invoke it. Lastly, Claude code is smart enough to install
00:23:51these skills. If you were to just copy this URL from GitHub, I, this is on the Claude code official
00:23:58GitHub. So I could copy this URL, paste into Claude code and say, Hey, I want to install the skill.
00:24:05Can you do it for me? And it will edit its settings on its own to add that. The last thing I'll talk
00:24:10about skills before we see this in action and then go into the skill creator skill is that skills can
00:24:15be both at the user level and the project level. Right now we are in the 35 test project, but what
00:24:21if I created another project and it was for, you know, my emails, it was my email project.
00:24:27Will that project have access to the Claude code skill? It depends. Normally. Yes. When I install
00:24:33something like the front end design skill, it's going to install it at the user level.
00:24:37You are the user, which means any project you, the user creates has access to these skills.
00:24:42We have the ability to have project level skills. So I could create or use a specific skill just for
00:24:48this project, just for our website, right? Maybe you don't want it to ever be used anywhere else.
00:24:53So just understand you have that flexibility. So let's actually put this to work. So I did
00:24:57forward slash front end design. I said, recreate the front end aesthetic, use the anthropics color
00:25:02palette, terracotta, et cetera, and give the card some visual weight, drop shadows, et cetera.
00:25:07Stay away from standard AI slop design. So let's see what it does this time with the skill.
00:25:14So here's what we got. And this is a huge leap forward from the first iteration. Now,
00:25:19is this perfect? Absolutely not. Does it still look AI generated? Yes. But look at the gap between the
00:25:26first iteration and the second iteration. And that's really thanks to one of the most simple
00:25:30tools out there. One of the most simple skills, which is the front end design skill. Now let's
00:25:34talk about the most powerful skill available to you inside of Cloud Code. And that is the skill
00:25:38creator skill, which is our next concept. The skill creator skill is not only good because it is
00:25:44trained to create high quality skills, but it allows us to modify and improve existing skills
00:25:49and measure skill performance. Over time, as you use cloud code more and more, you are going to find
00:25:55yourself doing the same things over and over. We all have a particular workflow. Anytime you identify
00:26:01some sort of workflow, some sort of repeated task, you want to turn it into a skill. And when we turn
00:26:06something into a skill and create a custom skill, we always want to use the skill creator tool,
00:26:11because it's actually going to tell us if it's even worth creating a skill in the first place.
00:26:16Its ability to measure skill performance is huge. It will automatically run AB test to see,
00:26:22is this skill even better than doing nothing at all? And it allows us to test skill improvement.
00:26:26So if you edit a skill, it will test it against the original iteration. And when I say test it,
00:26:31it runs multiple tests and gives us back actual quantifiable data. This is massive.
00:26:37In order to use the skill creator skill, just like the front end design skill, you can either
00:26:41copy this URL and bring in a cloud code, or you can head to the plugin marketplace, find the skill
00:26:46creator skill here, install it, reload your plugins, and you're all ready to go. Now let's talk about
00:26:51some external tools we can bring into cloud code with concept number 20 and 21, and that is MCPs
00:26:58and CLI tools. So MCPs are a way for us to connect cloud code to some sort of external program.
00:27:04Cloud code has an entire section of its documentation dedicated to MCP tools and lists
00:27:11the type of tool you can connect to, as well as the command you need to copy and paste into cloud code
00:27:16to install it. So we have things like Linear, Hugging Face, Cloudflare, Figma, Guru, Monday,
00:27:23Notion. The list goes on and on and on. And with an MCP server in between us and these programs
00:27:30like Notion, I can then just talk to cloud code in natural language and say, "Hey, do X, Y, and Z
00:27:37inside of Notion." And it will do it for us on our behalf, just like as if we were inside of Notion,
00:27:42executing those tasks manually. So to use these MCP tools requires a couple things. One, you need
00:27:48the specific command. Like I said, the most popular ones can be found here inside of the cloud code
00:27:52documentation, which you can find by either searching cloud code docs MCP or just asking
00:27:58cloud code to do it. And then you're just going to copy this command and then you're going to paste
00:28:02it inside the terminal. Now, the easier way to actually do this is to just say something like,
00:28:09"Set up the MCP server or Notion for me." Cloud code will do all this on your behalf. It will search
00:28:17the web to figure out what sort of commands it needs to run. And then it will come back to you
00:28:22with a list of things you need to do. Because for me to set up a Notion MCP or a PayPal or a Stripe
00:28:30or Superbase, I need to have credentials. So that means I need to go onto a website like Superbase
00:28:36and there's specific secret keys that I need to bring to cloud code or put into the settings
00:28:42folder. Now that sounds kind of complicated if you've never done that before, but guess who's an
00:28:47expert in these things? That's right, cloud code. Cloud code will literally walk you through it step
00:28:51by step to like where it says, "Click on this link, go to this page, copy and paste this." So cloud code
00:28:58will do all the heavy lifting for you when it comes to MCPs. That being said, MCPs are kind of falling
00:29:04by the wayside. Even though MCPs were hyped to the moon over the last year and a half or so,
00:29:09they are starting to be replaced by CLIs. A CLI tool is a command line interface tool. Just like
00:29:17cloud code lives in the terminal, these programs also live in the terminal. And because both cloud
00:29:23code and these programs live in the terminal, they are perfect for one another. Versus MCPs,
00:29:28MCPs have a lot of overhead, which means when you compare an MCP versus a CLI, the MCP tends to be
00:29:33slower and less token efficient. So for 90% of use cases, if there is a CLI available, you want to use
00:29:41that over the MCP. A great example of this is the Playwright MCP. So Playwright is a tool that if we
00:29:48connect cloud code to it, it will do browser automations for us. So it's like cloud code
00:29:52controlling the browser on our behalf, logging into websites, testing things. It's great. It has a CLI
00:29:57and an MCP. The CLI is way better. It's like 90% more token efficient. It's easier for cloud code
00:30:04to use. And it's a perfect case study for the CLI versus MCP dynamic. Now, how do we use CLIs? Well,
00:30:10you know exactly what I'm going to tell you. There are specific commands to use CLIs. Or what do we
00:30:15do? We go to cloud code and we say, set up the CLI for me. Now, things like CLIs and MCPs, if you
00:30:22aren't on bypass permissions, you didn't start cloud code with dangerously skipped permissions,
00:30:27it's going to ask you to do things that will probably tell you to open up a new terminal window
00:30:31and paste in these commands. If you're on bypass permissions on, it won't do all this for you.
00:30:35Now, part of that's kind of scary, right? Cloud code going in, downloading something on my computer,
00:30:40installing it, running it, frightening. However, if you're using a rather popular CLI, one that's
00:30:47definitely vetted by the community, you shouldn't be too afraid of this. Something like Playwright CLI
00:30:51is a great example. The question then becomes what CLIs should I be using? And the truth is,
00:30:57I can't tell you what CLIs to use. It's entirely use case dependent. Do you need browser automations?
00:31:03Maybe. Do you need something like Supabase for authentication to set up logins for your website?
00:31:08Maybe, maybe not. The trap here and a trap a lot of people get stuck in is they, after hearing about
00:31:15CLIs and MCPs, they go down the rabbit hole and they search up every single CLI they could possibly
00:31:20use, every MCP that might be relevant to them, install all of them and use them like crazy.
00:31:25Less is more when it comes to cloud code, at least in my opinion. So over time, you will become more
00:31:32exposed to some of these CLIs and MCPs and it will become rather obvious when it makes sense for you.
00:31:37If you are just getting started, I think it's enough to know that CLIs and MCPs exist. And more
00:31:44importantly, that we can have cloud code install and use them for us. And even in many cases,
00:31:50create skills that help cloud code use them. So if you just know that's out there, that's enough
00:31:56if you're in the beginning of your journey. If you are a bit more experienced, well then
00:32:00it's just a matter of using cloud code more and also asking cloud code if does an MCP exist for
00:32:06this use case? Does a CLI exist for a use case? Oh, by the way, no cloud code. Go use your web search
00:32:12and confirm that's the case because these things are changing all the time. Every single day there's
00:32:17a new CLI coming out. Every single day there's a new MCP going out. So I'm not going to sit here
00:32:21and give you, you know, here's the 30 you need to know, although I do have some content that
00:32:26hits on some of these things. Just understand it's out there and that cloud code can find it,
00:32:31install it and use it for you. These are the kind of like superpowers cloud code has.
00:32:36Because what I want to talk about for concept 22 is the idea of few shot prompting. What is few
00:32:43shot prompting? Few shot prompting is the idea that when I prompt cloud code, like with front end
00:32:50design, I don't just give it a prompt. I don't just give it a skill. I am in fact going to give it
00:32:55multiple examples of what I want it to do in order to better get the output I desire. And when we talk
00:33:02about something like front end design, we can do better than just screenshots. We can give it actual
00:33:07code, actual HTML that's living under the hood that cloud code can use to better get us something
00:33:14we're shooting for. So what do I mean by that? Well, you know, we originally told our webpage
00:33:17that we were kind of going for an anthropic color palette slash design. Well for one, I can take
00:33:23screenshots of what I'm going for. Let's add some of these cards. Let's get the footer, but we can
00:33:29look at the actual code. So if I do control U, what we have here is the HTML. So if I then copy
00:33:39all of this and bring that into cloud code, that is in essence, a form of few shot prompting where
00:33:45I have screenshots plus the code to better guide cloud code to get what I want. And this is perfect
00:33:50for front end design. So I said, can we get the front end to better match the anthropic aesthetic?
00:33:56Here is the HTML from their site plus screenshots. I pasted all that HTML was 2000 lines,
00:34:01and then I drag and dropped in the screenshots. So let's see how that works. And here's what it
00:34:06came back with. And it definitely has much more of a anthropic vibe in terms of the colors. Now,
00:34:13is it an exact one for one copy? Is there still a bunch of work to be done? Yes. Just like before
00:34:18with the front end design tool. But the point is by using stuff like few shot prompting, you have
00:34:23more control of the output. Anything we can do to get away from just prompt and pray and feel like
00:34:30we have some sort of agency in terms of the output. That's huge. And few shot prompting is one of the
00:34:37best ways to do that. But part of the problem of having cloud code code things when we ourselves
00:34:42aren't a coder is that we don't actually know if what cloud code is creating is correct or right.
00:34:48I mean, I can look at the output and judge it based on its final merits. But is what it created under
00:34:54the hood correct at all? It could be complete gibberish. And for most of us, we'd have no idea.
00:34:58Which brings us to our next concept, which is adversarial prompting and adversarial code review.
00:35:05Now, adversarial code review just means look at the code and tell me what's wrong. Now,
00:35:13when you do this, there's a few things you want to keep in mind. First is the idea that
00:35:18AI in general and Opus and Sonnet is part of this under this umbrella is they look kindly on
00:35:24their own code. Whatever they create, they think it's good. So if I tell, hey, look at the code and
00:35:29tell me what's wrong, it might come back with a few things, but by and large, it's going to think what
00:35:35it did is correct. And that shouldn't really be a surprise. So if I want to make sure that what I
00:35:39wrote was correct, there's kind of two options. The first is I can create another terminal and I can
00:35:46start up cloud code again. So I have cloud code in two terminals and I have a different session
00:35:51of cloud code. Take a look at the code we created inside of this code base. And I tell it to be
00:35:56adversarial. I tell it to like, look at it with a discerning eye. I tell it, hey, imagine you're
00:36:01some nerd on Reddit who hates AI and tell me what's wrong with this code base, right? That's the first
00:36:05way to do it. The second way is to use a different AI entirely to do the code review. And a great way
00:36:10to do this is using the codex plugin for cloud code. So open AI, AKA makers of chat GPT, makers
00:36:16of codex have a plugin for cloud code. So if you pay for chat GPT subscription, 20 bucks a month,
00:36:22you can use this plugin. In fact, you can actually use it for free as well, but I believe the model
00:36:26is downgraded, but we can install this just in the same way we install a skill and have codex review
00:36:34our code. And they have a specific command for adversarial review. So if you're someone who's
00:36:39like, I just don't believe what hope is the saying. And I want a second set of eyes that aren't also
00:36:44Opus. Well, codex plugin is perfect. How do you install it? Very simple. Take this URL, paste it
00:36:50into cloud code and say, Hey, I want to install codex plugin and it will run you through it. And
00:36:55you just would do forward slash codex review or forward slash codex adversarial review. And I think
00:37:00this is really important, especially if you're doing something more complicated, you know, we're
00:37:04doing something here with a front end, like a landing page. There's only so much it can screw
00:37:10up, but the more complex the project, the bigger the project, the more you're going to get out of
00:37:15something like an adversarial review. And this is great, especially, especially, especially if you
00:37:20don't come from a technical background, you don't know what it built. So at the very least you want
00:37:25someone who does know what's going on, AKA another model tell you if it was good or not. So now we're
00:37:32in section four, the power user section, everything we've talked about up until this point is something
00:37:37you should be striving to master within your first few weeks. If you can get comfortable with
00:37:41everything up until now, you will be well ahead of your average cloud code user. Everything we
00:37:46are now about to discuss though, this is more for the power user, for someone who's more experienced.
00:37:52You don't need to start implementing these things right away, but you do need to know they exist
00:37:55because as you become more experienced, you will eventually hit some walls. And some of the stuff
00:38:00you're going to be exposed to here can help you. So just know they're out there, know there's
00:38:04possibilities beyond everything that we've termed fundamental or core to cloud code. And the first
00:38:11two concepts we're going to hit are all about custom commands, custom slash commands. So when
00:38:15I do forward slash, as you've realized at this point is there are a bunch of commands that will
00:38:22prompt cloud code to do something, things like clear, right? We talked about that with context
00:38:26raw and starting a new session, but we can create our own. Oftentimes they are in the form of a
00:38:32skill. So you use the custom skill creator to create skills, which you invoke with a forward
00:38:38slash command. One of the favorite ones I use is YT-Pipeline. This is my YouTube custom skill
00:38:46research workflow because skills and cloud code don't just have to be simple things like a front
00:38:52end design text prompt that says, "Hey, when we create a front end design, make sure you do X, Y,
00:38:56and Z." Custom skills, custom slash commands, they can be workflows. They can be something that tells
00:39:01cloud code, "I want you to do A, then I want you to do B, then I want you to do C, I want you to do D."
00:39:06It becomes like its own automation. That's what my custom one does. And it uses sub skills. So it's a
00:39:13higher order skill that calls additional skills below it. And it uses things like CLIs. So your
00:39:18custom slash commands, your custom skills, your custom workflows can be pretty complicated,
00:39:22but they can save you a ton of time. Like I said, this uses the notebook LMPI. It goes onto YouTube.
00:39:28It's able to use the notebook LM actual application through cloud code to help me do research. And so
00:39:36whatever you do that has multiple steps that you do multiple times a day, you should be turning into
00:39:41a custom slash command. Custom commands can also be something like hooks. Hooks inside of cloud code
00:39:46tell cloud code to do something before a specific command or after a specific command. And these can
00:39:53be huge productivity boosts. My favorite hook is one that plays a noise every time it finishes a
00:40:00command. You probably haven't been able to hear it in this video, but anytime cloud code finishes
00:40:04a command, I get an audible sound. Now, why would you want that? Well, over time you'll realize,
00:40:10especially when you have multiple terminals up or it's running on a long task and you kind of just
00:40:15forget about it and you find yourself doom scrolling in between. It's great to have some
00:40:19sort of audible cue to say, Hey, I'm done. Let's get back to work. I would totally do that if I was
00:40:25you. And it's very simple to create these things. You just say, create a custom hook or create a hook
00:40:31that plays a noise plays a noise. I can't type when pod code finishes a task. That's simple. And you
00:40:39can have that hook be anything. Doesn't have to be a noise. It could send you an email for all you wanted
00:40:43every time it completes a task. Right? Cloud code is infinitely customizable and custom slash commands
00:40:48and custom hooks are a great way to leverage that. Now, the next concept we're going to talk about
00:40:53are all about sub agents and agent teams. So when you're working inside of cloud code at any time,
00:40:59it can autonomously on its own spawn sub agents to do tasks on its behalf. Let's imagine I was
00:41:07researching something about cloud code skills and I wanted to know, I wanted to go out to the web and
00:41:14find me information about cloud code skills. Well, it could spin up a sub agent. That's all it's doing
00:41:20is web search. So cloud code is essentially creating a second instance of cloud code. Just like you
00:41:26created a second terminal and it's telling that sub version of cloud code, Hey, go search the web for
00:41:32me. It could do that for any number of tasks it wants. And sometimes it will create multiple sub
00:41:38agents, do multiple things all at the same time. You as the human being don't really interact with
00:41:44these sub agents and they have a specific task. They complete the task and then they bring the
00:41:49information back to the main cloud code instance. Now you have the ability to create sub agents on
00:41:54your own, but for the most part, cloud code does this all automatically when it needs it. You just
00:41:58need to understand how it works theoretically because you're kind of wasting your time to
00:42:02create sub agents on your own. But this system can have some downsides. Let's say I told cloud code
00:42:06to create a more complicated webpage than the one we've been working on. And so it creates three sub
00:42:11agents on its own. It creates a sub agent for front end design. What we've been doing. It creates
00:42:16another sub agent for authentication, like a login page in it creates another sub agent for payments.
00:42:23Think like Stripe. The issue is these guys don't talk to one another. Front end design doesn't
00:42:28talk to authentication. Authentication doesn't talk to payments, which can be an issue because
00:42:32we're all trying to be on the same sheet of music. Yet with the standard sub agent, you know, paradigm,
00:42:39they're all in a silo. They do their own thing and they come back to the main cloud code session. And
00:42:44then this guy has to figure it out. If everything actually matches up well enter agent teams,
00:42:51agent teams is similar in that cloud code will spawn these sub agents. However, they now talk
00:42:58to one another. So front end design can talk to authentication. Authentication can talk with
00:43:06payments. Payments can talk with front end design and oftentimes it will actually spawn an additional
00:43:12sub agent whose entire job is to coordinate what these three guys are doing, right? Kind of acting
00:43:19like a middle manager for a team. It's trying to mimic a real human team setup, which means your
00:43:24primary cloud code doesn't have to coordinate everything on its own. Everything is done at the
00:43:29sub agent level. This is great. This is a more sophisticated setup than your normal sub agents.
00:43:35However, it's not for free because of that coordination to have all three of these guys
00:43:40talk to one another that costs more tokens. Furthermore, this is an experimental feature.
00:43:46What do I mean by experimental feature? I mean, you need to explicitly enable this
00:43:50inside of your settings.json. How do you do that? Well, you can go to dot cloud and edit yourself,
00:43:55or, and you know what I'm going to say, copy this page or copy the URL, go inside of cloud code and
00:44:01tell cloud code to enable agent teams on your behalf. Once you do that, to actually use agent
00:44:06teams, you need to explicitly tell cloud code. You want to use agent teams. In their example here,
00:44:12they said create an agent team. So it's not going to do it automatically. No matter how you say it,
00:44:17you have to say create an agent team. So let's see this in action. We're going to say create an
00:44:21agent team to work on this webpage, one agent for front end design, one agent to create a form
00:44:26submission for people who want to join our newsletter and one agent to research ideas for
00:44:30a blog machine. So again, you need to be explicit. You can see here the agent team it created. So it
00:44:35has a front end designer shows its task newsletter, dev task, blog, blog researcher task. And in this
00:44:43main cloud code session, you will get updates as the sub agents complete their tasks. And while
00:44:48they're working, let's talk about multiple sessions. Cause we talked about this for a second. And this
00:44:52is something you probably see all over the place, which is people with like nine terminals up at
00:44:57once, all working on the same project. This is something you could do. I can have pretty much as
00:45:02many terminal open. It has many terminals open as I want to work on cloud code. In a sense, I can
00:45:09mainly create my own agent teams. I could have one terminal here. That's working on front end design,
00:45:14one terminal here. That's working on the form, et cetera, et cetera. When it comes to these multiple
00:45:20sessions, there are some things you need to think about mainly kind of what we talked about with agent
00:45:24teams and sub agents at the beginning, which is like, are we having cloud code all work on the same
00:45:30piece of paper, changing different things at the same time. If you had three humans at a desk,
00:45:36all working on the same report and the report was on one notepad, you can see the issues you would
00:45:41run into. On top of that, I will say from personal experience, you can only have so many terminals
00:45:48working at the same time before you're just doing so much context switching that it feels like you're
00:45:54being productive. You're giving it a lot of prompts, but are you actually focused on any one particular
00:45:58task? Are you actually moving the needle? Are you actually moving forward in any meaningful way?
00:46:02I would argue after two terminals, maybe three, we're kind of just messing around at this point.
00:46:10If you see someone with nine terminals up, this is productivity theater. We're just pretending to do
00:46:15something. We're not really working at max capacity, but if you are going to go the multiple terminal
00:46:20route and you're going to work on a ton of things at once, you're going to want to deal with that
00:46:23problem that we talked about. Three people at desk working on one notepad, that's going to have issues.
00:46:27How can we deal with that? Well, one way we can deal with it is get work trees. So get work trees
00:46:34attempt to solve that problem. Without work trees, we have three people, three agents, three terminals
00:46:39working in the same project. They overwrite their work and we get a conflict. However, with work
00:46:44trees, instead of having three people at the same desk, everybody has their own desk and their own
00:46:50copy of the files. They do their work and eventually all the work gets consolidated. And guess who makes
00:46:57sure it all adds up when we put it together? Clod code does. Now to use work trees, you'll open clod
00:47:02code like normal, but you're going to use the work tree flag and then you're going to name that work
00:47:06tree. So something like front end design. So it'd be clod dash dash work tree, and then name the work
00:47:12tree. And then you would repeat that for all the additional terminals you open. And like you see here,
00:47:17this one is called main. This one is called front end. This one's called tests. You would work on
00:47:23those things in parallel. And then at the end, you would go back to your primary clod code session and
00:47:27say, take a look at these work trees. Let's merge these and make sure it's clean. A more advanced
00:47:33concept, especially for Git. So if you're just getting started, don't get confused and start
00:47:37using this right away. Now let's talk about another power user concept. And that is frameworks. Things
00:47:43like GSD, get shit done. Things like BMAT, to some extent, things like superpowers. It's this idea
00:47:49that there are GitHub repos out there that essentially act as an orchestration layer around
00:47:56clod code and change how clod code works at a somewhat fundamental level. It's still clod code.
00:48:01You're still in the terminal, but programs like this change how it plans certain things change the
00:48:07way it executes changes the way it records its progress. And the idea is that by using these
00:48:13orchestration layers, they help clod code do something. Most often it helps clod code handle complex
00:48:20tasks, complex projects. My take on this is that there's a lot of useful things in these
00:48:25orchestration layers, but you need to be somewhat experienced to understand when they make sense,
00:48:30because just like with CLIs and just like with MCPs, you can be a kid in the candy store and
00:48:35you think you need to use every single orchestration layer that is out there for the sake of it. As if
00:48:40you're just one orchestration layer away from really mastering clod code. I don't think that's
00:48:46the case, especially as time goes on. When GSD first came out, the way it handled things like
00:48:51context rot was pretty brilliant, especially compared to how clod code does it now.
00:48:55Yet a lot of the cool fundamental things of GSD and other frameworks like it have begun to be seen in
00:49:03the base level clod code. Like clod code out of the box does a lot of these things now. It's constantly
00:49:08being updated. It's constantly importing these best practices. That's not to say these frameworks
00:49:13don't work. It's to say, hold off on using them right off the bat. Less is usually more with clod
00:49:20code and just understand what you're actually using. Don't use these frameworks for the sake
00:49:26of them. Let's talk about the next concept, which is triggers and scheduled task, which can be kind
00:49:31of a confusing thing. Triggers, that can be a lot of different things. That's just when a certain
00:49:37thing happens, we do another thing, kind of like hooks and scheduled tasks are when we want clod
00:49:42code to do something on a schedule, even if we aren't there. When it comes to scheduled tasks,
00:49:47we have some options and this all kind of plays into triggers. And that's the idea of loop. So
00:49:53with the loop command, I can have clod code do something on an interval, whatever it is I want.
00:49:59Now there's a few caveats. First of all, it's session based. So I have to have this exact
00:50:03terminal open for this to work. Secondly, it has a timer. So max seven days as of posting this. And
00:50:10they originally came up with three days and now it's seven days. Point being you're somewhat limited
00:50:14because it has to have that terminal open. So if I shut my computer down, that loop ends.
00:50:18So something like a loop is great for, let's say you were constantly creating deployments
00:50:24and you were constantly updating your website and you wanted a loop that every 30 minutes
00:50:29checked deployment status. So almost like a mini skill in a sense, but it's hyper-focused.
00:50:37But what if you want to have clod code do something all the time, even if you aren't there?
00:50:41The answer is it kind of depends. So we talked earlier about the clod code desktop app. One
00:50:47of the things that clod code desktop app does pretty well is scheduled tasks. So if I go over
00:50:52here and hit scheduled, I can create a new scheduled task, which will run every single
00:50:58time, even if I'm not there. So I can do a remote task that will work in the cloud. So let's say,
00:51:04Hey, every morning at 7 AM, I want you to check my GitHub and do X, Y, or Z. It will do that.
00:51:09It can even do a local task. So I could have a local task for clod code searches. The web
00:51:14goes on YouTube uses the notebook LM tool and finds me the top 10 trending YouTube videos in a niche.
00:51:21It can do that as well. One of the caveats of doing that with the local task is that again,
00:51:27it's not session based. So the desktop app will create a new clod code session each time,
00:51:32kind of like creating a new terminal, but my computer needs to be on clod code needs
00:51:37to be open in some capacity. So understand that it's not exactly a scheduled task no matter what,
00:51:43but clod code does have the ability to create scheduled tasks on windows. So I have one that runs
00:51:48a GitHub API and pulls the trending get hubs for that day. Every single morning that has nothing
00:51:54to do with clod code windows, my computer, my machine itself does that, but clod code
00:51:59wrote the script. Now we just have a few concepts left. And one of them I want to talk about be
00:52:04alluded to at the beginning is ultra plan. What is ultra plan? Well, ultra plan is kind of like
00:52:11plan mode, but it's like a super plan mode and it works in the cloud. So if I do forward slash
00:52:16ultra plan and say, come up with a monetization plan for our app, it's going to say, Hey,
00:52:21do you want to run ultra plan? I'm gonna say yes. And then it's going to start a clod code session
00:52:26in the cloud. So it gives me a link and you can see clod code, just sort of like the chat app
00:52:31working here. Now there's a few differences with ultra plan. One it's not totally confirmed,
00:52:37but based on how it works, the assumption is there are multiple agents under the hood on this,
00:52:44working on this particular plan. So it's a little more powerful than your standard clod code plan.
00:52:48The second thing you'll see is that once it gives it back to us, it's a little easier to change it
00:52:54and give it sort of your input. So here's what it came back with an ultra plan. One of the differences
00:52:59is that I can just copy anything and I can leave a comment on that plan. I can even give it in emoji
00:53:06response because that's the best way to code, but overall it gives you a, just a nicer interface for
00:53:12it. That's the big difference. It's the fact that there's probably more powerful agents under the
00:53:16hood again, not confirmed. It gives you a different, more easy to use interface. And it's easier as
00:53:22well to give feedback on specific parts. If you approve the plan, you just click down here,
00:53:27approve the plan, and it reports it back in the terminal. And then you get the option to implement
00:53:31it within the session, start a new session, start a new session, just clears the conversation,
00:53:36but retains the plan. So we get a brand new context window or cancel. This just came out this week. So
00:53:42expect a lot of changes with the ultra plan. Now we've got two more concepts to cover in the second
00:53:46to last one is remote control. How you can use cloud code from your phone. The easiest ways with
00:53:52remote control. There's a few other ways, namely channels, but I will tell you for most people,
00:53:57remote control makes the most sense. You just use it from your cloud mobile app,
00:54:01but you need a session open. So the idea is I already have a particular session open.
00:54:05I'm already working on something inside of cloud code. And then I hop on my phone and I use cloud
00:54:11code remote with remote control. It's almost like I'm streaming my terminal. So it's pretty much,
00:54:18I see the exact same thing just on my phone. So it's useful in that sense. And if for whatever
00:54:24reason, your computer shuts down or it goes to sleep. If it comes back up, the session will
00:54:28reconnect on your phone as well. And the last thing I want to talk about is finding new tools
00:54:34for cloud code. I alluded to this a little bit when we talked about CLIs and MCPs and looking for them
00:54:39on GitHub via cloud code, but you can do this yourself. If you go to the trending session of
00:54:44GitHub, this is a great place to find up and coming tools for cloud code. Better yet you can have cloud
00:54:51code do this for you every single morning. Like I said before, cloud code gives me this and it breaks
00:54:55it down by section. So I see AI related GitHub repos. So you can see right here, this is the
00:55:01trending for today. Wednesday, I can see the top 10 repos that came out in the last week. I can see
00:55:07their stars. I can see their topics, descriptions. I can click inside of them. And this allows me to
00:55:13have a pretty good pulse on what's going on in the cloud code world. And if you're someone who's
00:55:19serious about this and always wants to be kind of on the bleeding edge of what's happening and what's
00:55:23trending, this is the best way to do it. Now, not every one of these is going to blow your mind away,
00:55:28but sometimes you'll see something very interesting and early, and you can also
00:55:32filter it by month. It doesn't have to be by day. So you can see all stuff, all sorts of great stuff
00:55:38here because it is pretty crazy. The stuff that's coming out every day, like it is every day, every
00:55:42single day, there's something new coming out. And instead of waiting for someone to sort of hand it
00:55:46to you and tell you, Hey, here's something you should try out. Why not just go to the source?
00:55:50And it's also really interesting to see what is trending because just by diving into these and you
00:55:56can point cloud code at it and say, Hey, clone this repo. Tell me what's going on. You're going to learn
00:56:00a lot too, because there's a reason these repos are blowing up. So that is where I'm going to leave you
00:56:05guys for today. I hope this video gave you some sort of framework, especially if you're just getting
00:56:11started, especially if you come from a non-technical background for where you need to start. When it
00:56:15comes to cloud code, this is an amazing tool yet. It is very intimidating. If you don't come from
00:56:21that sort of background, but I promise you, you can get through it. You just need to know where you're
00:56:26going. So as always, let me know what you thought. Make sure to check out chase AI+ if you want to
00:56:30get your hands on that masterclass. And other than that, I'll see you around.