talk to a model. With all that being said, let's really quickly cover how to set these different models up. I'm going to be using Codeex, Claude Code, and Anti-gravity. You don't need to know anything about any of these platforms in order to run these examples. And if you're already very familiar with...know, Claude Code and you've chosen to use that as your main agentic coding platform moving forward, you can skip over to the next section of the video. But I want to make sure that we all have an equal playing ground here. We all understand how each of these platforms work under the hood. So there are three
Поиск по транскрипциям
link to the tool and resources and notes from today as we go along. All right. So, you can see an example right here. So, you just have Claude code running. And the first thing you want to do is just make sure that you have Claude running. So, you can either go to terminal...Claude desktop, I assume, because Claude desktop has the code tab now as well. And then you're just going to type in Claude here. Make sure you've got Claude running. We're going to hit yes, continue. And now we can start copying and pasting these instructions directly into the terminal here and just seeing if it works
link to the tool and resources and notes from today as we go along. All right. So, you can see an example right here. So, you just have Claude code running. And the first thing you want to do is just make sure that you have Claude running. So, you can either go to terminal...Claude desktop, I assume, because Claude desktop has the code tab now as well. And then you're just going to type in Claude here. Make sure you've got Claude running. We're going to hit yes, continue. And now we can start copying and pasting these instructions directly into the terminal here and just seeing if it works
Cheers. Bye-bye. Today I'm going to show you how to turn Claude into a worker that never stops. It's called Ralph Claude Loop and it's nuts. This thing keeps coding until the job is actually done. No babysitting, no handholding, just results. It's brand new and developers are losing their minds over it. This...insane and I'm about to show you exactly why everyone is talking about this thing right now. So, what is Ralph Claude Loop? It's a plugin for Claude code that turns Claude into an autonomous worker. Instead of stopping after one try, it keeps going until the task is really done. Think about how Claude normally works
read all of the files in your Obsidian Vault, which is a folder of text files. But with the Obsidian CLI, it can give Claude code not only those files that it can read and access, but it can also give Claude code information about the inter relationships of those files. So you can see so cloud code...that oh this file is connected to this file and this file and that gets very interesting in terms of what claude code can understand about you and uh understand about uh all of the relationships between the things that you're working on. It can start to surface patterns about what you're thinking about that
just the best in terms of being able to code. It even started this whole vibe code revolution. When we actually take a look at what Claude's main offering is, coding is the number one use case that most people site. You can see one user here says Claude is so good at coding and reasoning. Is it just...head and shoulders above GPT4? Another user says Claude is a magnitude above GBT4 for coding and other tasks in my opinion. And this user claimed to have spent a,000 plus hours with LLMs last year. So now the reason this actually changes the game is because Google is actually the number one. And I understand that you might
thinking. While it's loading, I'll look at the comments here. I have not seen prompt cowboy. Should I look at it? How much better is Claude code? Okay. Replet uses Claude. Okay. Replet, Cursor, Bolt, Lovable, they all use Claude. They just have their own prompts that Claude then is forced to run through to give...result. And then they obviously have the wrapper around it for their UI. Um, but like Claude Code itself is just really good. I don't know how else to explain that to you other than it's good. Like it's like you saying why is LeBron better than uh I don't know somebody who sucks. Like
Claude Code is now a 247 AI employee, which means that it is always Cloud Code O'Olock. And that might be the nerdiest thing I've ever said. Anthropic has finally done it and launched scheduled tasks natively for Claude Code, which means every single process, every single skill, everything you've been building and using inside of Claude
going to do over here is I'm going to open up something called claude code. Right now a lot of people watching this will probably understand what cloud code is. If you don't basically this is a version of cloud that runs inside your terminal right as you can see right here. So we're going to click...continue. That's opened up all code. I actually have tons of training inside the AI success lab if you have never used Claude code before. So if you just go, this is a free community connects you with 43,900 members. If you go to here and then you type in Claude, you'll see that we have loads
description. So, that is this my system prompt. Then going to save. The next thing I'm going to do, I'm assuming you've already downloaded Claude Code. If not, you head over here to extensions, type, you know, in this case, Claude Code, but realistically, whatever model you want. Give that button a click, click install over here...least, you know, a $10 or $20 a month plan, you're good. I'm then going to go to the top right hand corner here, click this little claude code button, and now I'm just going to move back a bit and start asking it to help me. Now, what I want to do is I want
sent to Claude with every single prompt you make. And if you have the right rule set, you're going to get so much better performance out of Claude code. Over my hundreds and hundreds of hours of using Claude Code over the last year, these are the nine rules I use. I've put in where...long-term fixes that address the root cause. And again, it stresses simplicity, simplicity, so it doesn't overthink it and write way too much code. These nine rules have improved the performance of Claude by 10x. It really is amazing. So, make sure copy paste this, create a new claw. md file in your project, paste them
things you're going to need to make sure you do is to activate the artifacts tab so when you ask CLA to generate content like code Snippets text documents or website designs Claude will create an artifact that appears is in a dedicated window alongside your conversations if you already have this setting enabled then you're completely fine...what Claude is able to do in this react dashboard so I'm going to go ahead and just give it a couple of seconds and whilst Claude is coding away I can then even think about how in the future I could add this to certain presentations or what I would even want in other scenarios
been a little daunting to look at. So, I had my friend Vin and he clearly explains what Obsidian is, how to use it with Claude Code, how to set up these commands that really drive the most out of Claude and all the LLMs. And it's an incredible episode, like a really gamechanging episode because I think that...people who understand how to use Obsidian and how to use Claude Code together, they're going to be able to live happier, healthier, and wealthier lives. Why? because it gives incredible ideas to you on tap. So, I know that the people that stick around to the end of this episode, I think that for a lot of them
codeex for instance, I wouldn't call this a gemini. md. I'd call this an agents. mmd. If I was in claude code, I wouldn't call this an agents. mmd. I'd call this a claude. mmd. Whatever file you use here doesn't really change the idea. The idea is that at the very...dark mode and then it won't make that mistake again. This lets your knowledge accumulate over sessions. The first time that you use, you know, Gemini or Claude Code or Codeex or whatever, you know, you're only going to have, let's say, one rule or one preference stored. And so the number of errors that the model
same prompt. And we'll see which one performs better. a journey through existence says are locally hosted models finally going to surpass proprietary models for decoding tools like claude code I don't expect so the biggest problem is like this cla code is so powerful isn't it that's the thing it's so powerful...Look at that. All right. Even with sound effects, even though the sound effects were terrible, but you get the point. Like Gemini is still the goat. Claude Code, Claude Opus struggled. Deep Seek V3. 1 totally failed. and Gemini is still winning. All right, so don't worry about that peeps. However, if you want something open source, then
redid and rebuilt. Um, we upgraded from Vue 2 to Vue3 and Nux and um, it took two months and through about 75% of that time I used Claude Code and I had $1,100 bill and I was using Opus 4. 5. 4. 6 wasn't out yet um, during some of that portion and I said, you know...less flexibility and they're more way more for beginners, I guess you could say. Although I'm not even sure if I want to say that because claude code and those type of like AI agendic idees um that's the other category like claude code would be one cursor would be another one um those type of tools
brings back a much loved feature that's kind of been experimented with, taken out, put back in. Well, it's fully back. Ultra Think is totally back into Claude Code and you need to be using it. What Ultra Think is the ability to force Claude code to work extra hard on tough problems. And I've been using...they put it in like the rainbow colors. All you need to do in any of your messages is type the term Ultraink and it will force Claude Code to think very hard about what it's doing and put more effort in the creation. Hitting enter on that will start the process and it is going to make sure
then whatever your HTML block is, you just click on the blocks, grab that, hit custom HTML like so. Then you're going to go back to Claude, copy that code that you've just created the code with. Go back to the HTML block here. Paste that in. Hit publish. And then it might not work straight away...simple as that. Now, you know, the thing to note here is like some people say, "I'm not technical enough to build tools. " But, you know, Claude just writes all the code for you. So, you just describe what you want in plain English like I've done today. And then you've got the models to power
that it's not scale. The people who say that are not scalable people. Um, here's Claude's plan. An AI empires 2 inspired web interface for orchestrating claude code agents. Visualizing your kit code base in RTS where AI agents are units that can be spawned assign tax and monitored in real time. Agent types. Okay, we went...then I will continue my answer to your question, Luke. All right, looks good. How good, by the way, how good is this interface for a live stream using claude code like this? Now you can see everything really well. Feel like this is freaking great. Um, how do you market a web app, get an audience? You want
vibe coders are having their apps hacked and taken down because of bad security practices. Here's what I try to remember for everything I build in Claude Code. One, I start with planning. Two, I do execution. And then three is security checks. Right? So, every time you build a new feature in your app, you're planning, then...Claude just finished building a feature into your app, right? It went through the entire plan and finished the plan. You want to make sure every time Claude finishes a chunk of code or a feature, you run this security check that I'm about to show you here. And keep in mind, all these prompts I'll put down