Hermes Agent OS is INSANE! 🤯

Hermes AgentOS is the most powerful automation agent system available right now. Today I'm going to show you exactly how to use it to automate your entire business. I'm going to show you how to automate videos, content, SEO research, and a whole lot more all from one single system. And I'll show you the exact workflows I use every day that make this agent 10 times more powerful.

But here's the thing. There's one part of the setup. I want to show you later that almost nobody's doing it. When you see it, everything changes.

Your agent gets smarter. It knows who you are. It knows what you do, and it just gets better and better the more you use it. So stick with me all the way to the end because that one thing makes all the difference.

Let's get into it. I'm going to answer some of the questions on Hermes AgentOS, how to set up, what you can do with it, et cetera. So one of the first questions we had from the community here from Craig is how to do video editing with videos using something like Hermes, right? So how would I do that?

How would I approach it, et cetera. So there's a really good skill for video editing. It's called hyper frames. Inside Hermes agent.

So you can set up hyper frames of Hermes agent. What does that look like? Here's an example of what it looks like. And if you want to use it for video and stuff like that, you give the skill set up to Hermes agent and then it can use it, right?

So let's say for example, you wanted to edit a video quickly. What we can actually do here is we can, for example, go into Hermes, which you can see over here and we could say, okay, edit this video with hyper frames. And so let's say for example, we've got a video over here that we need to edit, right? And so what we can do here is we can grab that, get the file path, go back into Hermes and say, edit this with hyper frames.

Chad says, could Hermes help me set up and install this dashboard? Yeah, absolutely. So this dashboard that we have set up here for agent OS, you can actually grab the prompts that we use to build it inside the air profit boarding, as you can see right here. And you can also grab the zip file.

So we have an attached zip file inside the community here, and you can just upload that, give it to Hermes and say, Hey, go off and install this for me, right? Or set this up for me. And it would just work directly like that. So for editing videos, I would recommend using hyper frames.

Let me show an example of this in action. So we actually connected it to Haygen so we can generate AI avatar videos with Haygen and then we can edit them directly with something like hyper frames, right? Here's an example. So if we pull this up, this is a video we created and this video was fully edited and generated with hyper frames, right?

And so you can see how it basically took the avatar and turned it into a beautiful video. So it's a really powerful video agent. If you know how to use it, we have more training on that inside this section, as you can see, right? We have a full guide and video tutorial on how to do that.

So the workflow would be, you set up high frames at Hermes agent, you give the skill set up to Hermes agent, and then it can use it. You edit, you say, edit this video with the local file path. And then here's an example, full guide of how to do it. So another question from Karen here, which is what's the best Hermes and workspace launch setup, right?

And she said it's the fifth time she's starting over with an AI agent. So first thing I would say here is don't overcomplicate it. If you want an AI agent and you're happy with your existing setup, you don't need to start again, right? You can just pick one to save time.

So for example, if you're happy with Claude, you don't need to recreate another agent setup, right? You can just stick with Claude and learn that and master that. Now, if you want like the full setup on how to just install it from scratch with Hermes agent, with Hermes workspace, just follow the commands here, right? So like you can just follow the commands here, step-by-step.

There's a quick start guide here, and you can just go with the one line install. This is a good question. So Dan is asking about like the memory setup with Hermes agent, like how to really improve the memory setup. So for example, we have a full system with memory setup here using Obsidian and OMI, right?

So Dan is asking Omni on Windows or Linux. Is it worth it? Here's what I would say, right? Start with Obsidian first.

You don't need to set up OMI and just see if you're happy with the setup. Then from there, you can decide if you need to use OMI as well, or if it's just going to overcomplicate things for you and whether you actually use it day to day. So I always go with the idea of trying to keep things as simple as possible, and then just building up from there once I understand the concepts, because you can stack in as much as you want, but really what you want to understand is like the foundations first, especially if you're not sure about it, right? So start with the foundations, see if you're happy with Obsidian.

If it makes sense for you and it works for you, then you could be like, okay, I want to try OMI as well, which is great for, for example, tracking and building out new memories every single day. So for example, I can export these new memories to Obsidian daily and it's taking notes on me all day with OMI, right? But for you, if you're just getting started out, you could stick with Obsidian, keep it simple, and then build from there once you're comfortable with it. Also just going back to hyperframes, if you're using hyperframe, you're editing like local files, be really specific with the file path and what you want it to do and what edits you want it to make, right?

So you can see it working right there. Now I've got another question here, which is, has anyone tried Hermes and Reemotion? So Reemotion is really cool. It's hyperframes.

It's another sort of video skill that you can use inside Hermes, right? And so with Reemotion, essentially what you can do here is you can create animations, you can create videos, et cetera, using Hermes. And the way this works is if you want to install this and start using it to create videos using AI agents, you would grab the GitHub here, go back into Hermes agent and just install the setup. And by the way, we have this video from hyperframe set up now.

And so it will come back to you with the new file path once that's done. Now as Hermes is working on that, let me show you how you can use, for example, like Claude with Reemotion. And it's the same process, whatever agent you're using, this is how it works, right? So first of all, I think I've installed this already.

So I'm just going to ask we check back in a second. Now you can see an example of a video we've created with hyperframes right here using the same sort of project. If it doesn't show up, like it didn't for me inside the chat, then you can just say, okay, like open, why isn't it working? And then it will give you the commands to start running that with hyperframes.

Let's try something else now. So if we say, okay, use Reemotion to create a video about a ICO, just five seconds long for a quick example, make it fun plus interesting. And we'll send that also check open human as a request here. I've got a tutorial on that.

I don't use open human. I don't find it as useful as something like, for example, Hermes agent or Claude. I tend to find those are better. And a comment here, which says you demand Julian.

Thanks for staying on the frontier of these updates. Happy to help. Thank you. I use official says I was running Hermes agent on hugging face spaces, but the space is getting paused.

Yeah, that's why I run it locally. I just find I don't want to mess around with those setups too much. Now, if we have a look at Reemotion, so you can see an example of what the video you can create looks like here. So it's pretty simple to set up.

And basically the way this process works is you grab the details from GitHub, paste it into your AI agent to set up. And then from there you can use videos, for example, like this, right? So we said use Reemotion to create a video about a ICO, just make it five seconds long for a quick video, make it fun and interesting. And then you can see the video is saved here.

Let's open this up. So we've got the video, as you can see right here, it takes about five minutes to do. And that's my process. And so web says, Hey, do you know, I really like these types of videos.

You think Hermes can replace and it's an automations, 100%. 100. That's exactly what it's for. I think, right?

So the old way was that you would go into any 10 you'd mess around. Sometimes the automation didn't work. You have to be technical to understand HTTPS and everything else. And it was very messy, right?

Or as for example, with the new way, you can just give, let's say, for example, you were trying to create like a setup with any tenant and create videos. Let me show you an example of what this used to look like. So this was really like last year as a process. So the old way is like you have to set up everything like the API key and mess around with that.

It'd be a lot of work, right? The new way, if, for example, if you wanted to create like videos with AI is that you can just teach it, for example, the skill like Reemotion, it can create videos for you or you can give it grok 4.3 and then you can set up an agent OS like this where you can generate videos or images, whatever you want, right? And there's no coding required. There's no technical setup.

It doesn't take hours to set up and test and make sure it actually works. It just works like this step-by-step. It's really easy. Marcial says, what is your favorite local LLM for Hermes?

You've shown many. I actually, I don't use local LLMs that much. Practically even on a Mac studio, like I just don't think it's powerful enough to run them as efficiently as something like grok 4.3. And also if you want to generate videos and that sort of thing, like 99% of people's setups are not going to work.

Having said that the best one I've seen probably, I would say Quen 3.6 was okay, right? But just okay. And then Gemma 4 is super lightweight and very fast, but it's not that smart, right? And so I just, I feel like when you're using local LLMs with your agents, particularly Hermes, you don't get the same quality of responses, whereas you'll get much better responses just by using a free API, but you could go over to open router.

You can get something like our alpha, right? OWL alpha. And then you can use a free API. And you'll get much better responses than a local one.

So web says, Hey, thanks for the videos. Love it. I'm literally thinking into use Hermes agents for this example. Awesome.

Do you think Hermes can replace NA10? Yeah, 100%. I just covered that, but yeah, for sure. Oh, that's a good question.

So paradigm says, if it's my first day working with Hermes, where would you start? So I would literally give Hermes all the context you can, so it understands you as best as it can. And then from there, I would look at, okay, based on the information you have about me, what can you help me set up and automate, right? Because your use case is going to be different to mine, right?

So my sort of process for doing this is number one, give context to Hermes so they understand you as best it can. And that's why something like Obsidian and the memory system here is so useful, right? Because if I give a new, even if it's a new AI agent, if I give it all this information about me and who I am and what I do and the relationship between everything, as you can see, like it links together, then it's going to be way more understanding and way more practical, right? So you want to have something like Obsidian set up and then ask it based on what you know about me, what's the best automation I could set up today.

And just focus on one. Don't try and create loads, just focus on one. From there, implement it, refine it, test it, improve it. And if it works great next week, you can move on to another one.

And if you set up like one new automation per week, that saves you time and is actually personalized to your use case based on step one, then you can get, you can save a lot of time. You can get a lot of time back. Mr. Delight, cool name, right?

It says I use DeepSeek API. Is it wrong? Is it better than free alpha? No, I would stick with DeepSeek.

If you're happy with that, if it's giving you the results that you want, nothing wrong with that, right? Just stick with a lot of people like trying to overcomplicate things or try and switch APIs or switch tools all the time. Just focus on one. If you're happy with that, stick with that.

That's what I would say. So Dan was asking like, what's the best memory system? Here's what I use. So for me personally, I'm just going to stick to Obsidian and also OMI because it works so well.

Right now, when it comes to organizing your vault, right? So let's say, for example, you've got Obsidian as a vault set up from context and memory inside your AI agent. When it comes to setting this up, I would get one of your agents to clean it up and organize it. So for example, we go into Claude here.

I'll show you an example of how we clean this up and how we got it better. You can see that I trained it on examples of what I liked, examples of Obsidian vaults I really liked, and then he organized it based on that, right? And so your AI agents can connect locally to your Obsidian vault and organize it for you. I don't think you need an extra plugin or extra GitHub set up unless you really want to overcomplicate it.

And this way you just save a lot of time. And also your AI agents understand your Obsidian vault even better because it's the AI agent itself that's organized it and compartmentalized it into an easy system that any agent can understand. I think the best memory system that you can have is the simplest one. That's the easiest to understand and organize.

So if you limit the number of plugins and extensions you add to it, it's always going to be simpler and easier to use so you can get better outputs. This is a pretty cool question. So Faith was asking, can you build a WordPress website with AI? Right?

Here's what I do. Here's my process. And I'll show you some examples of websites we've built. You can do this with Hermes, you can do this with Claude, et cetera.

So let me show you some examples of analytics and sites that we're building with AI, right? So you can see, for example, this site is growing from like zero traffic all the way up to a peak of 71 clicks per day, right? Here's another one. So it grew from four clicks a day all the way up to a peak of 95 clicks a day.

And so these websites are ranking really well with AI SEO and we use agents to build them out. So how do we do this? I wouldn't use WordPress if you can avoid it simply because it's quite a bloated content management system and it's easier for AI agents to use something like Netlify instead. So what you can do is build out the website locally using something like Hermes or Claude, and then you can connect your personal access token from Netlify and deploy to a custom domain.

Let me show you an example, right? So if we go to Netlify here, these are all our websites, right? And so these get deployed by our agents using the agent OS system. So what I do here is we basically, for our websites, we will put it in a keyword here.

We'll plug in a case study that we want to create content with, right? And so that will deploy to five different websites that my AI agent has accessed it. Once he's done that with it, we can then see inside a history, what pages we've created and what keywords we've targeted, right? So for example, these five pages, literally what I did was paste in the case study of my own and gave it the keyword.

And then it went off, wrote the content and deployed it to this website using my AI agent inside the agent OS, right? And so it can create content, it can deploy, it can build, it can rank it. As you've seen with all my case studies of Google search console, you can see the websites are ranking pretty nicely. And that's basically what we do, right?

And the cool thing about that is anytime you have design changes or layout changes, if I'm like, okay, improve the navigation here or improve the setup or add more, for example, related blogs, et cetera, it can easily do that because the AI agent is connected to Netlify. So if it needs to make design changes, it can just add them at any time. So that's what I recommend instead. If you actually go to the classroom inside the airport for boarding, then you go to AI SEO automation.

As you can see right here, we have a full setup guide and we show you exactly how to set it up right directly here. The other cool thing with this is because my agents inside the OS have access to my memory, which you can see down here, it can write more personalized content, right? So it understands me, understands my business, understands everything about me, what I'm working on even. And so when it's writing the content, it takes that context and use it to write better content, right?

It's more unique. It's more personalized. It's more relevant. And then if we, for example, are targeting a keyword, like you can see right here, you can see that my content shows up here.

You can see my website's rank directly here, Reddit posts, et cetera. And so it works really well for the AI SEO. Do you use open spec? No, I've not heard of that before.

Be a hundred percent honest. Let's have a look at it. So this website says, can you explain how the memory in Hermes works? Tonight I'm creating my obsidian memory, but I see something like this.

Yeah. I don't really like the default memory systems inside Hermes. The reason for that is like, they're not, you can have a memory for free, right? So if you're using, for example, Honcho, I think there's actually like, it can, it has a limited number of credits and then you have to upgrade to get the rest of them, which I don't like, right.

Cause I can build a better system with obsidian directly here. And if I want to link my local obsidian vault to my AI SEO, I can do that. I can do that. So if I want to link my local obsidian vault to my AI agents, all I do is I would go into Hermes like this and I would just say, okay, find and use my local obsidian vault and use that as your memory, right?

It's literally as simple as that. It's so easy to plug in. I'll show you an example. So you can see the prompt here, find and use my local obsidian vault and use that as your memory.

That's way better than I think some of the options inside Hermes for memory from what I've seen so far. And you can see like it pulls in all that data and all that context right here. Like even for example, recent SEO articles, it can pull in that information from obsidian vault and then use that to write better articles in the future. That's how cool it works.

So I much prefer obsidian as a memory than any of the T4 options. Paradigm says, that was great. I just set up obsidian last week. Appreciate your jump.

Happy to help man. Thanks so much for joining. And John was asking about paperclip, right? Paperclip versus NA10.

So has anybody tried to replicate an NA10 workflow using paperclip? So I would keep it even simpler. If you want to replicate an NA10 workflow, just tell Hermes agent the workflow and ask it to replicate it. So let's say for example, you are trying to replicate an image workflow.

You can actually screenshot that workflow and then give it to Hermes agent and say, how can I set this up? Or how can you set this up for me without NA10? And the reason that I recommend this is like NA10 is kind of, it's almost like the old way of doing things, right? Because with NA10 you have to build the workflow manually and you're like the builder yourself.

With Hermes agent, when you give it workflow, for example, like when we create the video with Reemotion, Hermes is a no code alternative to NA10, right? It's doing the heavy lifting for us. It's doing all the technical stuff for us. And so we don't need to get NA10 involved.

And the thing to note here is if you want to build a massive team of AI agents, then paperclip can help. But if you just want like a simple NA10 automation that you want to move into your agent instead, then just give it the details and it can do it directly. You don't even need paperclip. By the way, if you want me to answer questions for you, like I'm doing for everyone in the community, link in the comments description if you want to join or it's available at theairprofitboarding.com.

So I think that's it for all the questions. We've answered everything here. Let's see what questions we've got inside the community too. So Brian says, I didn't like Hermes when I first started using it, but tweaking it more and more today and it seems like it's getting better.

Yeah. The cool thing about Hermes is it has a self learning loop, right? So it always gets better over time. So the more you use it, the better it gets, the better it gets, the more useful it becomes.

What do you think of Hermes instead of NA10? 100%, right? 100%. When it comes to usage limits, you're only limited in terms of usage limits with Hermes by the API you're using.

So if you don't have a limited API, if you're using something, for example, like step 3.5 flash, which is free or our alpha, which is also free, then you can use these free APIs and there's no limits, right? There's no usage problems there. So that's what I'd recommend. And then, like I've mentioned, and NA10, it feels to me like an older version of AI versus something like just using Hermes directly, right?

Hermes is like the fastest, simple way. Let me show you an example. So what a lot of people do, particularly with any sort of system is they have point A, which is where they are and they have point B where they want to be, right? Where they want to go.

And so what a lot of people do is they do this, right? Especially with NA10 because you set up an API, that takes time. Then you have to figure out how to configure NA10. Then you also have to figure out how to use HTTPS.

Then you have to test a workflow. It works great, but then it stops working. Maybe your API runs out and then eventually you get to your target, right? But it might take like way longer than it should.

Now, for example, if we have a look at AgentOS and we look at where we're set up ReMotion here to generate this video, right? We just went from point A to point B, just like that, right? Boom. So it was way faster and simpler using Hermes agent than it was using something like NA10.

NA10, it might take me a few hours to set up. Hermes agent, it takes me a few minutes. That's the difference. There's a platform called co-founder.

Can you check it out? Might check it out in the future. Is it possible to connect cloud design with Hermes agent? That's a great question.

I think what you would do there, this is what I would do, right? I would actually get open design on GitHub and I would plug that into Hermes agent, right? And then you can get Hermes agent to use computer use or you can use an MCP between them to use each other, right? That's the way that I would look at it.

And also, yeah, that's the way to do it. Or you can just design with Hermes directly. You don't need to use cloud design or anything like that. Josias says, good to see you here.

Thanks for joining. So Paperclip is designed for this. I'd recommend using Paperclip. It's a pretty good system for building like a full AI agent company.

It is a little bit complicated to set up, but if you wanted to build like a full AI agent company with no humans, you can do it with Paperclip and it connects all your agents. Yes. So it says Hermes comes with a lot of skills installed. Yeah.

You can use open design instead. Can use Anthropic with Hermes? No. As far as I'm aware, you can't, but what you can do, this is what I do is inside my agent OS I have Claude and then I have Hermes over here, right?

And so they can link together and speak to each other. And also I can use them in the same setup, right? And that way I can use a free API with Hermes and I can use my Claude subscription, right? Which is pretty simple and easy.

Yeah. So Marciel says maybe open design instead of cloud design will work better because it's open source. Absolutely. I think it would integrate better as well.

So that is basically it on how to use Hermes agent to automate anything. We've covered a lot of stuff. If you want to get the full agent OS system that I've shown you today, you can get that inside the AI profit boardroom. Just go to the new daily tutorials and you can see the setup video guide and the full prompts and even the zip file for setting this up, which you can see down here, right?

So you can get in the resources section and you can just give that to your agent and say, build this, right? This is my AI community that helps you save time and grow and learn with AI automation inside the community. You can ask questions, actually answer this stuff with videos as you can see daily. And I give you the TLDR if you don't want to watch a video so that you get your questions answered daily with super useful personalized tutorials inside the classroom.

You can see all of my new trainings on this stuff. Inside the calendar you can jump on four weekly coaching calls and inside the map you can connect with people in your local city as you can see right here. So this is all inside here. And also some people watching this might be like AI agents are too complicated and that sort of thing, right?

So you can see here, for example, like all the testimonials here, we have 158 pages of testimonials and wins from people learning growing with this stuff. If you can type a sentence, you can use this stuff, right? You've seen how simple and easy it is for me today. And if you want to use AI agents in the same way, we have loads of wins.

As you can see, we've helped a lot of people. It doesn't just work for me. All right. Thanks for watching.

Hope to see you inside the AI Profitable Audio. Link in the comments description or go to theaiprofitableaudio.com. Thanks for watching. See you on the next one.

Cheers. Bye bye.