Do you want to see something cool? Just look at this. I'm in my terminal and I typed out this command. Now watch what happens. It fires up this tool called Claude Squad. By the name, you can probably guess what it does. It generates multiple versions of Claude code in a single terminal. You can assign different tasks to each instance, and it's honestly pretty amazing. Next, let me show you how to use it. The guide is pretty clear. You just press N to open a new instance of clawed code. Let's say this one will work on the front end. So, I've named this the front-end session. Then let's go ahead and create another one. Again, I'm pressing N and this one is going to handle the back end. Now I have another instance where the clawed code agent is focused on back-end tasks. This tool is honestly amazing. I recently stumbled upon it while scrolling through X. So now let me show you how it works, the kind of cool stuff you can do with it, and how Claude Code and Cursor are now working together. If you like the content we're making, I'd really appreciate it if you could subscribe. Right now, we're also testing out memberships to support the channel. We've only launched the first tier and for now it offers priority comment replies, but subscribing helps us see how many of you are interested and want to support what we're doing. Claude Code has done something wild. They've integrated Claude Code directly into your IDE. The feature is called IDE integration and it lets you use Claude Code inside your IDE. Don't just get excited that Claude Code is now available in your IDE. Focus on the facts. What does it offer and what are the real benefits? Normally, if you open the same project directory in both Claude Code and Cursor, they're already working on the same code. But this integration adds extra features on top of that. First, the obvious one. You can now use Claude code directly inside cursor. Here's something I found really useful in this mode, selection context. Whenever you highlight something in the code, it's automatically shared with cursor. If you copy something, cursor's agent already has it. No need to paste it manually. Same thing happens here, too. You just select something in the IDE and it's automatically shared with clawed code which is honestly a solid feature. Another thing that made cursor really powerful is how it automatically shared small errors like lint errors and syntax errors with the agent. I'm not sure about console errors yet, but lint and syntax errors are already covered. Claude code is now going to do the same thing. Next, a quick recap in case you don't know how this works. Here's the process. First, install claude code. Just copy the command, paste it, and it'll install in your terminal. If you're having trouble or don't want to use the terminal, copy the command, open cursor, and ask it to do it for you. It's really not difficult anymore. Inside cursor, open your integrated terminal. That's the one built into cursor. The first time you do this, it'll install an extension called Claude Code. Let me show you. This is what gets installed when you connect the first time. After that, if you want to run it inside Cursor, open the command pallet, search for run clawed code, and hit enter. It'll open the claude code terminal in a new window right inside cursor. You can close the old terminal if you want. Now you've got the claw code terminal embedded inside cursor and it supports all the commands we've talked about. Now here's a major benefit. Claude code doesn't limit your base model's context. Right now I'm using claw force on it. If I use the same model in cursor, cursor limits the context per model usually to save API costs. But this is the raw model so you get way more context. That alone is a huge advantage. And Claude also manages context extremely well. Let me show you something. I'll type a prompt. Make a website in HTML, CSS, and JS for a fishing business. When I press enter, you'll see it generates to-dos, which is incredibly useful. You might remember our channel has showcased a bunch of videos using cursor in a structured way. Claude Code does that by default, and that's one of the biggest reasons to use this tool. Even if you don't like working in the terminal, maybe you prefer an IDE, because let's be honest, the terminal can scare a lot of Vibe coders. This setup makes it really easy for them to still use clawed code. Also, one more thing, it shows you how many tokens are being used in real time as you work. That's pretty helpful, too. About the tool claqued in the beginning, I tried using the same method with that tool. So, let me show you. I've just initialized Claude Squad here. And now, let me open up this session. This is the command for checking the IDE connection. You can also see the other commands. These are the commands for Claude code. I've opened a Claude code instance using this through Claude Squad. As you know, Claude Squad allows you to run multiple instances of Claude code. When I run this command, it says no available idees have been detected. So, clearly, there's some issue. It's not able to detect the IDE. Honestly, this only blocks us from accessing a few features I mentioned earlier, and they're not super critical. The multi- aent capability that Claude Squad offers is way cooler. If you know how to work with these agents together, it becomes much more useful, way faster, and just a lot more powerful. Here's a classic example of how to use multiple instances of these cla code terminals. If you go on X and search multiple claude code, people have been doing some crazy amazing things with it. Let me give you a simple example that came to my mind. So you have a front end and backend app. They're connected by APIs. API blueprints can be created to define exactly what needs to be built and then the front end and back end are structured around those APIs. That's exactly what I asked one of the instances to do. I gave it the spec and told it to start creating the front end based on that. And if I check this now, you can see it's still working. Now I go into the next session. And if I open it, I can prompt it to do something else. This is just crazy. It really shows how amazing this tool is. I stumbled across it and thought I'd share it with you. These tools are coming together in a way where you can mix and match having multiple agents working on different parts of a project. Sure, it's going to take a lot of planning to build a full-scale production ready app, but if you use all of them smartly, it'll be way faster. The trade-off is that it's going to be a lot more expensive. That brings me to something Claude recently introduced. It's called the Max Plan. It costs $100 per month, and what it gives you is a much more generous usage limit. Basically, an unlimited token cap that resets every 5 hours. It's not like the older pro plan where you had a monthly quota. And of course, the free plan doesn't offer any serious token allowance. But with the max plan, if you connect it to Claude code, you get unlimited tokens with some of the models. Even with the best model, Claude Opus, they've implemented an hourly rate limit instead of monthly limits, which is a lot better overall. It also works with OpenAI codeex, which is another Claude code style coding agent that runs directly in your terminal. If you like this tool and want to install it, let me show you how. Installing it really isn't that difficult. on their main website and I'll leave the link in the description below. Just head over and visit their GitHub. All the install instructions I'm about to walk you through are listed there so you can copy them directly. The first things you need to have on your system are two tools, T-Max and GH. You need to have both installed. Right now, T-Mox is only available for Mac OS and various Linux distributions. So, unfortunately, this won't work on Windows. Let's hope the developers gain enough traction to release a compatible version for Windows as well. The commands to install T-Max are listed right there. If you go into the GH directory, you'll find the command to install the GitHub CLI too. After that, they provide an automated installation script. Just copy it. It's right there and paste it into your terminal to run that will install the entire tool for you. Once it's installed, you can initialize it using the command CS, which stands for Claude Squad. As you can see, it's been initialized here. This is a session I created and it's now active. But you can't initialize this tool in just any directory. To use it, you need to be in a folder where GitHub is already properly set up. So if you have an older project that you now want to integrate Claude Squad into with multiple agents working on different tasks and Git is already initialized, then you're good to go. If not, you'll need to run a few commands to set that up. Use these commands to initialize a GitHub repository and make your first commit so the tool can detect your working tree. The working tree is the folder where you make code changes and as you create different versions of your codebase, those changes get tracked. Claude Squad and the Claude powered code without a Git repository initialized, it won't function. And for those of you unfamiliar with using the terminal, if you already have a project, just paste the commands. If you're starting from scratch in an empty folder, you'll need to create a simple file to make your first commit. You can use this command. It creates a file called ignore.txt, txt, which is just a placeholder. Then initialize your GitHub repository using the earlier command. And finally, use the Claude Squad command to launch the tool. And there you have it. That brings us to the end of this video. If you'd like to support the channel and help us keep making tutorials like this, you can do so by using the super thanks button below. As always, thank you for watching and I'll see you in the next one.