Agent OS: Build ANYTHING with Hermes + Claude!

This new AgentOS gives Claude, Hermes, and OpenClaude AI super powers. So right now, most people are using AI completely wrong. And today I'm going to show you AgentOS, a full operating system that makes your AI agents way more powerful, and I'll even show you how to build it for free. Plus why this changes everything.

Think about your phone. Without an operating system, every app would just sit there doing nothing. They couldn't talk to each other. They couldn't share information.

You'd have to do everything manually yourself. That's exactly what most people's AI setup looks like right now. Now, AgentOS connects all of them, one place, one screen. And suddenly your AI goes from a bunch of separate tools to one incredible system working together for you.

With Claude, OpenClaude, Hermes, all operating together, all in one place. And it makes every single one of your AI agents way more powerful. But here's the part that gives you real superpowers. Your agents actually remember everything.

Your goals, your plans, your business. The more you use it, the smarter it gets and the more work it automates for you. I built this in one session. I'm going to walk you through exactly how I did it step by step.

So you can build your own today. And by the end of this video, you're going to have your own personal AI operating system up and running today. And stick with me to the end because there's one system I'm going to show you. And that's the part that most people miss.

And it's the single biggest unlock you can get with AI agents right now. Let's get into it. Today, we're going to be looking at AgentOS, which is a new system I've set up for managing Claude, OpenClaude, Hermes, and any other agent that you want to manage directly. So you can see, for example, here, this is basically a whole operating system.

This is not just like talking to Claude directly or for example, Hermes or OpenClaude. What you can see over here is we have all of our agents ready to go inside this operating system. And then additionally, we can chat with them. We can see our history.

We've got a full status of all of our agents on the dashboard here. And then also we have this section here where it is recording into our memory, our goals, our daily journals, our memory system. We can actually see every single chat that's been logged over here across all of our agents. And it's a super powerful way to give your AI agents superpowers, right?

Because you can basically manage everything that you need to from one single place across all of your AI agents. I'm going to show you exactly how I built this directly and how you can do the same today. For example, you can see here that I ticked off creating the agentic OS project, as you can see. And then we can set more goals.

We can track them. It tracks how many goals we've done, how many are active, how many completed. And we can chat to, for example, Claude here. So for example, I've said, are you working?

It says, yes, I'm working. How can I help you? We can do the same with Hermes, with OpenClaude, et cetera. We can switch between them.

We also have a control room for every single agent, which is pretty amazing. So you can see here, it manages all of our API keys, all of our auth providers that we can see our previous session history, which is pretty cool. We've got our skills. So these are all the skills that we've actually set up.

So for example, if you're in stuff like Paperclip or Claudico, we've got all of that here. Plugins, we've got the Kanban linked as well. So we can add new tasks from the chat into Kanban too. And then we've got insights as well.

So over here, we actually have analytics of Hermes over the last 30 days. So for example, how many sessions, how many tool calls, how many tokens we've used, what models we've used the most, et cetera, and it breaks it all down beautifully, right? It's just a really easy way to manage your agents inside an amazing platform. It even shows you the activity patterns here.

So for example, we've got Monday, Tuesday, Wednesday, when we used it the most, the peak hours, the active days, notable sessions as well, even breaks down the longest session, the most messages is a really powerful way to just manage your agents in one place and have an amazing sort of operating system. Now the difference between this and going directly into terminal here is huge, right? If we just go directly into terminal here, let's have a look, look at the difference, this is terminal user base, which I think 99% people do not want to use, it's just, it's too complicated. It doesn't feel that nice to use.

And if you compare it side by side versus this, this mission control is actually way better. I'm actually giving it away inside the AI profit board. And if you want the full operating system that I've built, you can just direct message me once you're in the AI profit board and I'll send it over to you. And what's really cool about this is I literally built it in one session with Claude.

So let me show you an example of how we did this inside Claude. So this is Claude desktop, right? And create this new chat and I said, create a beautiful operating system hosted locally for managing Claude for a website connected to Claude should be like a beautiful mission control dashboard. Then allow me to control my open Claude, my Hermes, any other agents and separate systems inside the dashboard.

It should be a mind blowing operating system command center for managing Claude and my agents. And it was like, how should the dashboard talk to Claude? So I said, Claude CLI bridge. So for example, if we go inside Claude over here, we click on this and we're like, okay, can you build out an SEO blog for me?

And then we can just hit enter there. You can start thinking here. We can also talk to it as well. So it's got a microphone option as well.

And you can see how Claude just replies to us directly inside the chat and we can see if it's thinking or not. It has a few questions first before it's going to get back to us. Now, I also gave it the documentation, build this with AI agents. So you can see here, Hermes and OpenClaude are set up there.

And then it gives us ideas on what you want the stack and feel. So I just went with the recommended, which is Next.js and Tailwind, right? You don't need to worry about that. Like all this stuff you can do just by texting Claude like a friend inside Claude's desktop, and then it can build out everything for you.

Right? So you don't really need to worry too much about like how technical this is to build or anything, because you're not coding anything. You're just texting Claude like a friend here. It says, I have a few questions before I start building.

Let me ask a couple of questions. What do you want me to build? How do you want me to build it? Blah, blah, blah.

Right. And so this is all stored locally as well, which is awesome because we have a full command palette that's linked directly in this live for our agents. Now, what I've also got is I've got a memory system, which is connected to my Obsidian vault, right? Now this is pretty amazing.

So you can see here, this has all of my documentation for Obsidian. If you're not familiar with Obsidian, basically it's a really powerful way to organize all of your notes. As you can see right here, we've got like this beautiful diagram where it's all focused on me. It's all built around me personally.

And then I can give this memory to any AI agent that I have. And it basically has full context of me, my life, who I am, what are my goals, et cetera, right? And so it's a really powerful way to create some awesome context with your AI agents, right? Now, the amazing thing about UNC here, for example, it's got all my social media channels.

It's got like all my information about all the agents that I use and my website, even like people that work for me. It's documented notes about who works for me and that sort of thing, which is pretty amazing. So it's got Goldie agency over here, which is my SEO agency. And then it links out to everyone inside my team, right?

Which is pretty insane. It takes notes on my team who I'm working with, et cetera. And this all revolves around me. It's a full knowledge base organized right here.

Absolutely amazing, right? Even like my notes about my grandma and stuff like that, which I didn't even know what like music I like. It's got like the music that I like and the index, Muay Thai, et cetera, right? All my personal, everything it has about me, right?

And this all stored with Obsidian. I have a full system inside the AI Profitable for this too. It's called the Infinite Context Engine. It's a full lesson about how to do this.

But basically what's happening here is Omi is recording my screen, recording my microphone all day, taking notes on me, who I am, what I do, et cetera. So you can see 40, 48 minutes ago, user is developing an agentic OS in their Obsidian vault, and then it creates tasks for me. And what it actually does from here is it links out to my notes. So I can export this to Obsidian every day.

And you can see here, it's taken 1,261 memories about me, right? So this is all stored inside my Obsidian vault. That's all linked together beautifully. And then we can empower our agents with my memory system inside Obsidian to basically go off and have full context about me, who I am, and anytime I interact with it, it works both ways, right?

So basically if I interact with Hermes, it will log everything into Obsidian, right? So you can see here, every exchange also saves to agentic OS memories inside my Obsidian vault, right? Same with flawed, same with open claw. Pretty amazing.

Now, at the same time, it can pull in the context from Obsidian. So all of the context, all of the notes, all of the useful stuff I have inside Obsidian, which is absolutely massive at this point, and I only set this up like last month, so I'd recommend that you do it. The earlier you start on this, the better, right? The better you can set this up, the more powerful your agents will be because they understand you, who you are, et cetera.

Let's see what we got inside the comments. If you have any questions as you go along, feel free to ask. So Anthony says, incredible work. Thank you very much, Anthony, for watching.

And Brian says, nice. Yeah, thanks. Yeah, it's pretty amazing. So for example, I can say to Hermes, okay, based on my Obsidian vault, give me some ideas on what I should automate today.

And then Hermes is now going to start thinking and start working for us, right? Now, in the meantime, let's also go through here. So we've talked about the Obsidian vault and the memory. We've talked about the agentic OS with our AI agents.

We've talked about why you would do that. It's just not even comparable to the terminal, right? And also showing you how to set this up directly here. Now you can see here, it says, based on the Obsidian vault, it's going to come back to us in a second on that.

Now you can see here, it's looking through my memories, looking for my ideas, et cetera. It's hosted it locally, which is awesome. So I can just use this kind of like a website. I can bookmark it and come back to it later, which is great.

And also because it's hosted locally, you don't want to host this sort of stuff on the cloud when it has so much information about you, who you are, et cetera. Now, if we go back over here, this is pretty cool. So it's. looked at my vaults across Goldie agency, they are profitable in the community, my Hermes work, et cetera.

And it's here's where your automation would move the needle today. Right. And so here's some stuff you could do. Here's what you can set up.

Here's how you could grow, et cetera. Here's some quick wins for you as well. And so it's offered ideas that are actual, very useful, beautiful ideas I can implement today. Right.

I could just take one of those and say, just go off and do this, right. And it will just build it for me locally in the same way that I've built this whole agentic OS, and that's the power of agents really is really a level of, and bear in mind, like 99% of people have never used AI properly, right. Especially in this sort of context. So let alone building an agentic OS.

So if you understand how to use and build this stuff and how to operate it, like you are ahead of everyone, everyone. Imagine how powerful you could be. If you've got three AI agents working in parallel like this, they will have context of you, your business, and they know exactly what to automate every single day. Amazing.

Right. Absolutely amazing. Let's see what we've got here. Trad says such a neat graph.

I've got to say like the obsidian vault is looking beautiful, isn't it? It's looking absolutely amazing. It's insane here. It's just got so much useful information about me, who I am, who I'm working with, who my team is, how it can help me, et cetera, right.

And so it makes it far easier for agents to understand how they can help me too. Now, also what we built in is a persistent sidebar. So it's got the mission control. We can stack in and layer in more updates here, and I'm using Opus 4.7 to build this.

So the cool thing about this is if I want to stack anything else in, I can just add in here. For example, I could add an SEO agent for the AI Profit Board community that could post my websites every day. And then I just go inside this section here and just add it. And like, that will be super easy to stack in.

I might do a video on that later. And then what we've also done here is added some more features, right? Now, bear in mind, like this was just one hour ago that we've done this. So it's so easy to build out.

We said, make it more modern, beautiful, clean, improve the UI, et cetera. Add a microphone section where I can type in my voice, make sure I can chat with Hermes agent and chat with OpenClaw too. Make sure you log every memory into Obsidian, have a goal section and a journal section here as well. Right.

Just, just the fact that your AI agents could see your journals, what you're working on, what you're struggling with every day is going to be magical too. So it's really cool. So we've got the journal here where basically you can scroll through, add new entries, explain what happened. You can see the first journal entry right here as well, and it stores it by date.

It stores all the recent entries as well. And then your agents get access to that as well directly. So if you go back inside the chat, once you've stored the new memory, guess what? It's going to be even more useful.

It's going to have even more context. It's going to be even more better. Right. Yeah.

And then we've set up a, an export option as well. So if you're in the AI profitable community, you can just get access to this stuff and I'll share it with you, which is pretty cool. And then you can go from there. Right.

Also it helped me fix OpenClaw, which was great. We can add multiple profiles and a finished setup nicely. Right. Really easy to use.

The other cool thing about this is for example, we can use Hermes inside here for free, right? Because we're using the Hermes agent setup, which is a free open source project. Right. So we can use that for free and it's connected to news portal using step 3.5 flash, which is free to use as an API.

So the API is free on the agent. The agent itself is free and you don't have to use Claude to build all of this. You could use Hermes to build it too. I just use Claude cause I knew it'd be the fastest and easiest and the least buggiest way, but you could actually build it for free with something like Hermes directly as well.

And then you've got the goals and everything else. So let's talk about like why we've done this, how it helps, what the whole setup is. Right. So basically this is an operating system that can run your whole life.

Right. Most people using it, like most people using AI, like a calculator, this is something which gives you a full control to automate anything with your operating system. Now, most people say, what even is an operating system? Your phone has iOS or Android, your computer has Mac OS or windows, right?

These operating systems will make every app, every tool, every notification work together in one place. Without an operating system, you just have a bunch of apps that can't talk to each other. That wouldn't be great. Right.

But that's what most people's AI setup looks like right now. It's just a bunch of disconnected tools. It's like chat GPT in one tab, Claude in another tab, some random AI tool they saw online in a third tab, right? There's no memory between them.

There's no coordination. There's no system. Agentic OS, which is what we've built over here, fixes that. So it's a personal AI operating system that runs locally on your machine and gives you one dashboard with a mission control to see and control everything.

So every agent, every memory, every signal is all in one place. Bear in mind, we don't have to just use Hermes OpenCore and Claude over here. We could, for example, add or stack in any other AI agent that we want to. Inside here, right?

This is just an example, a quick example of that, right? We could put any single AI agent inside here, whether that's, for example, yesterday I showed OpenHuman. You could probably stack that inside here too, if you just gave the details to Claude to get it set up, right? I'm calling this the Goldie mission stack, which is the new paradigm for running your whole business and life with AI essentially, so that you stop being the worker inside your business.

You become the operator running the system that does the work, right? And it's a three layer model built around agentic OS. So each layer here has one job, each layer connects to the others and together they make you unfairly powerful, right? So let's talk about this.

Number one is the intelligence layer. So we have Claude AI and Claude code working over here. And this is your thinking layer, right? Claude is wired directly into your agentic OS as a live connection, right?

It has full tool access. It has MCPs connected. It can write and execute code. It can read files, analyze documents, plan strategies.

In your mission control dashboard, you can see Claude status live, where it's online or running, and you can click open control room and you'll see a direct access to Claude over here, right? You can see, and you can see the full activity system over here as well. So you can see all the events and that sort of thing inside this dashboard. So there's no separate tabs.

There's no copying and pasting. It's just one click from your dashboard, which is beautiful. And you can see the Claude is online here. Herm is online here as well.

And then you've got the execution layer. So you've got tools like OpenClaude. OpenClaude is your local agent gateway. It's the engine that can route tasks between your agents.

It can manage sessions. It can manage multi-agent coordination. It's what makes everything talk to each other. Think of it like the router in your house.

So your phone, your laptop, your TV, they all connect to the internet through the router. OpenClaude is basically the router for your agents here. Inside mission control, you can see exactly how many agents are running, how many sessions are live, and whether the gateway is healthy. And then you have the research layer.

So you've got, for example, Hermes that can help you with this. So Hermes agent is a powerful tool that can run tool calls. It can implement Kanban style task lists. It can use skills and plugins to go out into the world and do real things.

If you need to research a competitor or run a multi-step workflow on a schedule, Hermes can do it, right? And in mission control, you can see Hermes model as provider and connect directly to it inside the chat here, as you can see, and then we have layer number four, which is the self layer. And this is grounded in Obsidian, right? And this is the layer most people never build.

And it's probably the most powerful layer, right? Because whatever tools come out in the future, it doesn't really matter. Obsidian is still going to be relevant, right? Like it's still going to have, it's still going to be useful, no matter what context you're in.

The earlier you build your second brain, the better. That's what I would say here, right? It's just a really powerful system. If you want to get more details on that, go inside the classroom over here and then go to the infinite context engine, right?

You can see the infinite context engine here, and this is how to give your AI agents a memory that never forgets, right? It's a free stack where you can use Hermes, OMI, and Obsidian to create an infinite memory. And we've got a full video tutorial on that right here. So this is the layer that most people don't build.

And this is really cool because your agentic OS isn't just connected to your AI tools. It's connected to you personally, right? It's wired into your Obsidian vault, into your personal knowledge base. So this means your agents know your goals, they know your journal entries, they know your notes, your ideas, your decisions.

They have context about who you are and what you're working towards. And there's three components that can live here, right? You've got your goals, so you can set your targets inside agentic OS and the system tracks them, right? So you see your percentage fill bars, you complete them, your agents know what you're working towards.

And this is super useful for your agents to understand too. Then you've got your journals, your daily entries. That could be voice, it could be text, it could be one file per day stored directly into your vault, right? So your agents can reference your journal to understand your current state, focus, and priorities.

And then you've got the memory, right? So every single chat is auto locked with a full vault search. So you can search across 1000 or thousands of notes in your memory and your entire conversation history. Basically your AI never forgets anything that you've told it because it's got the full context and that's inside the memory section here.

As you can see, it's all these notes. We can see them over here and we can see the full OMI notes. We can see all the recent ones. Everything is just inside here.

It's pretty amazing. Trad says, just got Cloud Code today and it's pretty nice. I seem to be a heavy token user. Yeah.

If you use it a lot, you will, but from what you can build with this stuff, it's pretty amazing, right? It's pretty amazing stuff. So when all four layers are running, you sit above them all. You are the mission operator.

You set direction, you check the dashboard, you approve results, you scale what works. You stop doing the tasks and you just start running the system that does the tasks, right? And the difference between someone using AI and someone running an AI agent operating system is the difference between having a hammer and running a construction company, right? Totally different.

That is a different. Now, some people say, I don't need an operating system. I just use AI tools when I need them. The right belief here is that's exactly why you're not getting the results you want, right?

Disconnected tools give you disconnected results. A system gives you compounding results, right? The agentic OS system, this is day one, right? Imagine what day 30 looks like.

It's pretty wild. Everything optimized and automated. It's what you need. And then every day your agentic operating system runs, it gets more useful because it knows more about your business, your goals, your voice, right, and a tool you open and close never builds on itself, right?

ChatGPT is not going to build on itself, but something like this will. Some people say building an operating system sounds incredibly technical and complicated. The dashboard you're looking at right now was built to be navigated like a phone, right? You see the agents over here, you click the one you want, you open it in control room, you work, and the complexity is hidden underneath, right?

Your job is just to be the operator, not the engineer. If you need any changes, you can just go back to Claude over here and you can tell it exactly what you want. That's it. As simple as that.

Other people say AI doesn't need to know about me or my goals. It just needs to answer questions. That for me personally is a limited belief, right? The right belief here is like the single biggest unlock in AI performance is convex.

When your AI knows your goals, your journal, your notes, your history, it output, its outputs go from generic to genuinely useful, right? And the self layer in agentic OS is what separates a toy setup from a professional one, right? Adding that memory, adding that obsidian layer is what gives you superpowers, right? Because it understands you, your business, understands everything that you're working on, even to the point where it understands what music you like and everything else, and it can just map out everything and understand your whole life, which is what we've done here.

Pretty amazing. Ricky says, what's your thoughts on running Hermes on a local versus VPS? I personally prefer local setups. I just think like all the issues that came out with OpenCore came from a VPS, right?

Now I do think they can work either way. I just prefer to run it locally as well, because I have more control over it. And also it can then link to the memory easier as well. Fun fact says some awesome projects to look into.

Oh, awesome. Oh, nice. All right, cool. So that's basically it for the agentic operating system, but super powerful stuff, so much cool things that you can do with it, it's pretty much endless what we could stack into it as well.

The goals, the journal, anything that you work on day to day, I would add it over here. So for example, for me, I could add like an SEO agent that can create content, deploy blogs for me, et cetera. And the other thing I would say here is most AI tools run in the cloud, but that means your data travels to somewhere else, right? And then it comes back to you.

The agentic OS runs locally. So your agents live on your machine. Your bot lives on your machine. Your sessions run on your machine.

This is super useful. Now you could connect it to the cloud, but the thing is, it's going to be more reliable, private and faster. And also something like this, you could never build with a cloud based system. I think it's much better to run it locally, right?

So that's basically it. We actually have over 100 prompts that you can use for this. As you can see inside the guide from today, we also have a 30 day roadmap, right? So if you want a 30 day agentic OS roadmap to go from zero to fully operational mission control, you can get that inside the AI profit board and community, but it's a pretty amazing system.

So just to recap what agentic OS is. It's a local first personal AI operating system. It runs on your machine. It gives you one dashboard, mission control, see the status of every agent, every memory, every signal.

Connects to Claude, OpenClaude, Hermes agent in one coordinated system. And you could add any other AI agent that you want inside there too. It grounds the AI in your personal vault. So your goals, your journal, your memory.

It turns you from someone using AI tools into someone running that AI operating system. Totally different thing. We've talked about the Goldie mission stack. So four layers, you've got the intelligence, which is Claude, your execution, which is OpenClaude, your research, which is Hermes agent and orchestration, and then you have self, which is the Obsidian vault, right?

And your role is just a mission operator. You set the direction, you review results, you scale what works. And yeah, just in terms of some people will say tools don't need an operating system, but disconnected tools give disconnected results, as I mentioned. Some people say it was too technical.

You literally just click all the complexity or the code is hidden. AI, some people say AI doesn't need to know about me, but context is probably the single biggest performance unlock you can get with AI agents. Others will say I already have Claude, but one tool is not a system, right? Going into Claude desktop, that is not an operating system.

That's just Claude desktop on a tool that's separate from everything else. This all links together. Of people say I need a developer, actually Claude code can handle all the technical stuff. As you've seen today, that's how we built it in literally one hour.

And then AI has no memory, some people say, but your vault has thousands of notes and that can grow over time. Others will say, I need to be watching it 24 seven. Actually your agents can run whilst you sleep. Multiple agents are overwhelming.

Some people say, but mission control gives you one screen, right? It organizes that. So some people say I've got shiny object syndrome or I struggle with overwhelmed, blah, blah, blah. But if you can manage everything in one place, wouldn't that be simpler and easier?

Absolutely. Right. So that's basically it. And if you want 115 prompt across every layer and the 30 day roadmap, you can get it all inside our community.

Let me put that inside the classroom for you. So we add it inside the new daily updates and advanced tutorial section, and you can get all the notes. Plus we have video tutorials and all this sort of stuff too. We also have a Hermes agent operating system and every single day, we're just adding new tutorials, new updates based on everything that's come out.

So if you want to go deep and learn really advanced stuff, this is great. But if you're just a beginner with AI, I would recommend the beginner to expert AI automation course over here. If you want to learn how to get more clients with AI, we've got an agency course. If you want to learn like the best automations I personally use, you can get them inside here.

And we also have an amazing community of 3000 members. So you can ask questions, get help and support. And you can see that it's really active. Lots of cool questions getting posted inside there.

Inside the calendar, you can jump on four weekly coaching calls inside the AI Profit Boarding. And you can also connect with people in your local city from the map, right? So you have this map where you can find people in your local city, connect with them, find other cool people using Hermes, OpenCloud and agents as well. So it's a great community where you can learn and we can all grow together.

Plus you can connect with me personally inside there. If you have any questions, feel free to ask. Otherwise I'm going to run through some of the other questions inside the community. This is something that I do just to help people and really just help people learn and grow.

So let's see what we got here. So Patrick just joined us. You can see that he was really interested in Hermes. So welcome Patrick.

By the way, you can see how welcoming the community is. It's just an awesome group of people helping each other and welcoming each other in, which I love. So let's look at the next question. So Ash says, I had to rethink my Hermes agent architecture.

And he actually talks about how he didn't have an operator, right? He had a very polite approval clerk. And he said, the new direction is treating Hermes less like a tool launcher and more like an operator loop. As you can see right here, we've got some other questions about Hermes.

And so Karen was asking, as I'm progressing on my Hermes setup, I have questions about these three topics. Number one, Hermes workspace. I saw this somewhere that we no longer need to install Hermes agent separately from Hermes workspace. Is this true?

I already had Hermes installed and then I gave it the docs installs workspace and it connected to the backend. Yeah, that's totally fine. Right? So you want to make sure that you have Hermes installed first before you use Hermes workspace, and then you can use them together.

So you have to install Hermes first. Then you install Hermes workspace and that's how you set them up together. So have a look at the next question. Model.

So I have Codex OAuth connected to Hermes agent. Chose that model, but also have open router for fallback with Olama and Confuso. Should I set everything in open router so that the agent can choose their models based on the tasks? Am I going to have duplicate setups?

So I would have GPT 5.5 as the main model and then have an open router as a fallback is totally fine. That's totally fine. And then you can have Olama as the last backup model in case everything else doesn't work. And then also for security, Karen says, my Hermes workspace didn't ask for a password to connect.

Not sure why this is interesting. So you can ask Hermes directly to set up a password for Hermes workspace if you prefer, but it's a personal preference for you. Right. And then number four, also with setting up OMI and also yeah, with Obsidian, this is interesting.

So bear in mind, like with OMI, it's recording my microphone. It records my screen, et cetera. Right. Which is useful.

But at the same time, if you're not comfortable using something like OMI, then you can limit its permissions so that it doesn't track certain things. The other option that you have is you can just get your agents like Claude and Hermes, et cetera, to automatically update Obsidian every single day with their latest notes and memory on you, and that way you don't even need to use OMI. You can just have your agents write directly to Obsidian if you prefer. This is interesting.

So Mike says Claude for small business. So he says he's just set up a, an AI consulting business and Anthropic comes with Claude for SMBs, right? But he says, does anyone else believe that this has stopped solopreneurs trying to use AI to help other companies? Here's what I would say.

99% of people, especially small business owners, don't have the time to log in or understand how to use Claude. And they probably never will. They might understand how to use AI or chat cheaper tea, but they're not going to have the time to really customize and set that up properly for their business. So there's always going to be business owners out there who could do that themselves, but just don't want to, and would prefer to find someone else who can do that for them.

In the same way that, for example, like we could all create social media content, but as a business owner, you'd rather hire someone because that saves us time and we don't have time to do everything. So I think the opportunities here are wide open, especially when you're such an advanced user, you can pretty much help anyone save time. If you understand how to use AI properly. So have a look here.

So Lara is asking about Hostinger with OpenClaw, but Hostinger, and she says she's trying to find the easiest way to run this for free, right? So if you want to use OpenClaw for free, you can connect it to our alpha on OpenRooter and that way you've got the free setup from OpenClaw and it's free to use. with an API like our alpha on OpenRouter. So if we go over to OpenRouter and we look for models and then we type in our alpha, you can see this as a free API, right?

That you can use here directly. So that way you've got the free API with OpenCore and you've got the free setup of OpenCore itself, which is free to run locally anyway. This is an interesting one. Um, so Craig says how to show people that you're saving them time, right?

So let's say for example, you've got a potential client and they're like, okay, I'm interested in AI, but I don't know how this is going to help me. Here's how I would focus on it. So you can ask them how much is their time worth? And then from there you can give them ideas on like how much each of the automations you offer actually saves them.

So for example, if I save two hours with an automation, it's a no brainer to set up, especially when that's two hours per week, which leads to a hour save per month. You can also ask them like, how much time do you normally spend on this task? And how liberating would it feel to actually be able to automate that task? This is like a standard sort of discovery question when you're speaking to a client on a new call.

The other thing you can do is set up a calculator. So I'll show you an example and I'll quickly go on to Claude and show you examples of how to set this up. All right. So here's an example, right?

So you can run through a calc, you can easily build this in one prompt with a prompt like this, right? And so you can say, okay, so your time is around here, as you can see, and that's per week, right? And so if we automate this, you get this back, right? Which leads to this, right?

And then you can clearly show them the results and you can show them how it works, et cetera. I actually share that with you as an example, in case you want to use it. And so that's four different ways that you can demonstrate this. Now, if you're struggling getting them on a phone call in the first place, then I would just do more marketing.

So for example, like if we want to generate more SEO strategy sessions, typically what we'll do is create more videos around SEO and then have the CTA inside the video to book a free SEO strategy session, we've actually got training on that, let's find this over here. So we've got a full step-by-step course on this over here. Let's see what else we've got here on the questions. Settle says, Hey, good to see you.

So look, so Trad says, I was building an AI automation, dealt my mom's business and ended up coding some unique tools. They say maybe 10 hours a week. Now that's insane. Congrats.

And says the AI couldn't handle the task on its own. So I had to make software it could use. Nice. So that's amazing.

And that's like, we're really early here. Imagine what you can do in the future, right? Once you learn how to do that, imagine once you've replicated it five or 10 times, right? It's pretty amazing.

Joshua says with the new features and setups is the Hermes course and they are profitable and getting updated. So the fundamentals are the same, but we also have new daily updates here. If you want to learn new cool stuff, but I would say if you set up any of this stuff up, just set up one thing at a time each week, just so that you don't overwhelm yourself, you know, get distracted with shiny object syndromes. And then also I'd recommend that when you're setting up new systems with Hermes, et cetera, you focus only what's absolutely necessary because you don't want to waste time setting up stuff or playing with stuff that you don't actually need or use in the first place.

This is cool. So Vybav asks, is it possible to use Hermes as a man in the middle, right, as an AI agent in the middle? So Vybav says, I was thinking if it's possible to use Hermes agent as a man in the middle, basically have an app with its own MCP, and I want to give the users an option to chat with it and get their work done through the MCP. But given that most of my users don't have cloud code, I was wondering if I could add Hermes as a man in the middle.

Here's how I'm thinking, right? The app logs in, chats are given a session, Hermes uses the MCP and responds back to the app. This is interesting. So I think what you could do is you could set up something like the agent OS, but you have it hosted on the cloud.

So other people can access it. The other option is that you could have like a cloud hosted version of Hermes agent, and then people can just talk to it through there. So for example, if you look at Max Hermes is a good example of this. Max Hermes, you can use Hermes in the cloud.

That has, for example, something like the API from minimax plugged in, but you could plug in like the cloud API, for example, or you could link this to a cloud MCP, right? Bear in mind, like if you wanted as well, you can actually link Hermes MCP to your app. So you can run Hermes as an MCP server. And that can, for example, be used by people using the app and then people can operate whatever you want via this Hermes MCP.

So that's three different ways, right? You could set up something like the agentic OS, but have it hosted on the cloud. The other option is you have a cloud hosted version of the Hermes agent. People just talk to it through there, or you can actually link Hermes MCP to your app using the MCP documentation right there.

I think that's pretty much it for all the questions. So I've answered everything, answered all the questions inside the chat. That's my full system for using agentic OS. Like I said, all the training is inside the app offer boardroom.

Hope to see you inside there. If you want me to answer your questions in video format as well, like this, feel free to join the community. It's just about helping each other, raising each other up and just learning and growing together, which is the goal. Plus you can get the agentic OS system inside here as well.

Yeah, pretty amazing stuff. See you on the next one. Cheers. Bye bye.