NEW Claude Agentic OS is INSANE!

**Julian Goldie** (0:00)
Today, what I'm going to show you is a Claude agent operating system. So what we've done is build out a mission control dashboard with all of our best AI agents and workflows plugged into one powerful system. We've got, for example, a mixture of agents built in over here so that we can create and preview anything that we build. We have, for example, Hermes Oracle that pulls in the latest news, and then we can create SEO content with it. We can draft social media content in one single click, and it links directly to the trending topic, as you can see right here, plus it organized and categorized it every day, every 24 hours on a rotation by what is the most trending topic. Then we have a voice-activated agent over here. We have a pipeline where we can go from ideas to implementation. So what I want to show you today is what I'm doing to create all of this and how I'm basically building out an agent operating system empire, and then also what we're using around this that's even more exciting. So, for example, what we do here is we basically build new things into the agent operating system on a day-to-day basis. Now, if you're wondering what is an agent operating system, it is one place where you can orchestrate all of your agents, all of your CLIs, and also every single custom workflow that you use. So, for example, if we generate images or thumbnails, then I could generate them in one click over here, right? If I need to generate, for example, AI avatar videos, I can do that in one click over here. If I want to automate SEO, I can do it all in one click over here, right? So my point here is that everything that costs you time, you can now automate using systems like this, and everything is just one click away whenever you need it. That is the power of an agent operating system. That is so much different than me going into Claude and being like, where do I click or where do I go? Where was that conversation last week about open SEO? I can't find it, right? So that's why we don't use Claude, and instead we use the agent operating system. You might say as well, okay, well, doesn't this cost tokens? No, because you can use free APIs, you can use, for example, very cheap APIs, you can use CLIs you already have plugged into the system, and also, for example, you can use local models. So however you want to do it, you can make it very cheap if not free, or use minimal tokens using this system as well.
Then the other point that I get sometimes is like, why not just use Claude directly? But Claude is not going to give you these custom workflow, it's not open source, right? So we can't customize Claude in the same way that we customize an agent operating system.
This looks and feels exactly how I want it to look and feel, because I've designed it literally from scratch. And if I want to delete anything or hide anything, I can on this side in this setting too, right? That is so much different than, for example, if you go into Claude and you're like, oh, I have to use and stick with their AI, I have to use and stick with their layout, I have to use their UI, you have no control, you have no control. Then the other problem as well, of course, is like, Fable 5 goes down, good luck building some cool stuff for having all your automation since like Claude made, right? You kind of lost.
You might also say, well, also like why do we need all this stuff? You don't, you just swap and change, you need to delete what you want. So that's the power of this stuff, and that's why we're doing it. I think it's probably the best system that I've seen out there for building this sort of things. The other thing that I like about all of this is that we can build for speed. So, for example, as soon as Hermes mixture of agents came out, which was like 48 hours ago, we built it into the system, and we built loads of crazy stuff with it. In fact, we even use Claude when we're testing all this stuff with our agent OS to document the whole system inside a beautiful website like this, so that people can just see and understand exactly what we've built, why we think it's the best, how it compares versus other models. We do the same, for example, with the compare section here so you can compare all of your models, you can see all of our local systems, but the point is that everything that we test, we automatically document because we're using AI to help us along the way. So the cool thing about this is if you're doing this, if you start doing this, if you're into AI and you want to get clients, for example, well, guess who clients are going to hire? They're going to hire the people who test the most stuff. How do you test the most stuff? You use an agent operating system like this so that you can automate every single part of the process, but also you can build amazing stuff that impresses clients, so they're like, oh, I want to hire that guy.