Hermes AI Agent: Build & Automate ANYTHING!

Today I'm going to show you how to build and automate anything with Hermes, which just hit 106,000 GitHub stars. And today I'm going to break down exactly step by step all the things that I know about it, how to use it, how to automate anything, how to build anything with it. This is by far the most powerful AI agent I've ever used and super simple and easy to set up, right? The other cool thing that I like about this is it's very smooth. It works and it's a lot more reliable than something like OpenClaw and it's free to get started, right? And also show you multiple different ways you can run this directly. So you can see here that we've actually got it running inside our terminal. So, you know, if I was like, okay, build a new website for the airprofitboarding.com, I could just go straight into the chat and say, build a new website for this site, right? And that's basically how we can get it working directly inside our terminal. Now there's multiple different ways you can connect to it, but I'm going to guide you from the basics and show you how to get it set up step by step, right? So, number one, how do you actually use Hermes? How do you get started? Well, you can go to Hermes Agent. It's available on GitHub. It's completely free. And this is an open source project, right? It was actually created by Noose Research, which are one of the most respected AI agent labs in the world. Everything that I would say about using Hermes over something like OpenClaw is that OpenClaw is a lot less reliable, breaks more. We've all been there. If you've used OpenClaw, AI agents like that, they break, they're not that smooth, they can be quite difficult to set up, etc. Hermes fixes a lot of those problems, right? So if you want to get started with this, let me break down exactly how you can get started. So essentially what you would do here is you want to do the quick install, right? So you're going to grab this terminal command and you just go inside your terminal like this and just install it as you can see, right? And hit enter inside the terminal. You don't have to open up your terminal. You just press command and space and open up the terminal there. The other option that you have, if you're already using something like CloudCode, you can just go inside CloudCode. You can copy this URL for the GitHub. You can paste in all the information from GitHub, right? And then you can say, install this for me, right? And you can just get it set up from there. And that's basically how you can start using it, right? How you can get free access to it directly. And now you can see it's going off and just beginning to build out our project for us. All right. So you can see here, for example, it's just beginning to build out the website and start creating it for us, right? Now you might be wondering, okay, what sort of use cases are there for this? How can you use it, etc? So some of the best ways that you can use Hermes AI agent is basically just to look at what you do day to day and then figure out, you know, reverse engineer from there. Okay. What do I do day to day and then how can I automate that? So for example, it could be something like, you know, creating social media content. It could be, for example, using this for web research and that sort of thing, right? If you actually want a bunch of skills and ideas for using Hermes, you can actually see that right here on this GitHub. So this is an open source project where they've shared all the use cases for, you know, ways people are using this, right? So for example, there's all sorts of plugins here, all sorts of skills you can use this for. But essentially what I would do is say, okay, you know, knowing what you know about me, how can you help me, right? Something like that. And you can do that inside the chat here. And it will just go off and start working with you. Now you might be wondering as well, okay, like you can see here, it's actually come up with a list of ideas in terms of what we can do and how we can do it and that sort of thing. Right now. So one thing you're going to note here is like, if you want this to be more personalized for you, here's what I'd recommend. There's something called Obsidian and you can link Obsidian to your AI agents, right? So for example, here, I've got details of like what we sell over here. We have details on who I am, et cetera. Over here, we have details on my business and that sort of thing, right? Even personal memories and that sort of thing. So what I could do inside Hermes is I can say, okay, go to Obsidian locally and use that as your memory. Right? So that's an option. You can also do something called Hermes memory and you can change your memory with Hermes right here. But basically what you want to do is really train and explain to your AI agent, okay, who you are, what you do, et cetera. Now, when it comes to memory and I'm going to paste in all these use cases here as examples. So you can see here now that it's looked at my Obsidian vault, which is just a simple free program that basically has all of my details. Right? So it knows who I am, what I do, has all my memories, et cetera. As you can see right here, super simple and easy to get that set up. And so what it will do is pull in from Obsidian all the notes that I have and then figure out how we can use this. Right? And all the best use cases personally for me, which makes it 10 times better. So you want to set up a memory. Now, my recommendation equals Obsidian, as I've mentioned, but here's something that's going to help you make this even better. Right? So when you're using something like, for example, Obsidian, Obsidian requires lots of notes. Right? Lots. So how do you how do you automate that? Right? How can you automate a memory about you and what you do, et cetera? Well, here's what I'd recommend. There's something called OMI. You see an example right here. And basically what this does is it records my screen. It records my microphone. It has access to what I do all day and how I use it. Understands, for example, it's got access to my microphone here. You can see like it reviews and sees what I do. It also has lists of memories and that sort of thing. It has a list of all my tasks. And then also we can rewind and see what I do. I see screenshots of what we've done and it connects to my apps. Now, one of the apps that connects to is Obsidian. So what that means essentially is that we can sync our memories from OMI. Let's take in all of my details here and then plugs it directly into Obsidian automatically. And so I don't have to do anything. OMI just records my screen, understands what I'm doing all day, and then uses that to generate my memories. So OMI plus Obsidian is amazing. Now, obviously, not everyone wants to record their screen and record their microphone all day and that sort of thing, which I totally understand. So you set your own limits with that. But for me, I can then export my memories from OMI to Obsidian plus sync them with Hermes, right? And more context, again, equals better results, right? This is what it's all about. So just to recap here, in terms of how to use Hermes to build an automated end fin, first of all, you need to set it up, right? How do you set it up? You can get the GitHub here. You can go to your terminal, run this command, run Hermes, follow the onboarding process. If you need useful commands, you can check that out here. If you want, for example, non-technical way, just use code code plus paste in the GitHub documentation and ask it to set this up for you, right? That's basically it. Simple as that. All right. So that's the setup covered in section one. Now, step number two is use cases, right? How are you going to use this? Again, you want it to have context on you. You want it to understand what you do day to day, and then you can use that to help you automate whatever you want, right? So for example, for me, it could be like content marketing, it could be AI tool development, business operations, right? Community management as well, and research analysis. Like it could handle any of those. And the key question you want to ask it here is knowing what you know about me, how can you help me, right? What can you automate? Now at this point you're going, well, it doesn't have information about me, so how do I make sure it knows stuff about me, right? And that's how you go to the memory section. That's why you're going to use the memory. So inside the memory, you want to make sure your AI agent knows you, cause better outputs, more context, better results. My recommendation is Obsidian. Obsidian requires too much note taking, equals super boring, plus hard work, right? So how do you organize that? Well, you can use OMI plus Obsidian, right? And then you export your memories from OMI to Obsidian, sync them with Hermes, equals more context and better results. How to sync them with Hermes? All you do is you go to, you say, go to Obsidian locally and use that as your memory. That's basically it. So go to Obsidian locally, use that as your memory. That is a prompt you want. And that's going to give you ultimate memory. So let's run through this one next as well. All right. So if you go to Hermes inside the terminal here, we can use something called Hermes paperclip adapter, right? And this is basically a way to run, you know, an AI agent team together using paperclip with Hermes. Now there is paperclip directly with the GitHub, but this is specifically designed for Hermes, right? And so we can grab that URL over here and we can say, okay, set this up. And I'll just do it inside my main AI agent. So we'd say run this plus set it up locally, right? So we can now run paperclip and that will have a setup with agent teams as well. And it will just take the GitHub details and start setting up too, right? Now there's one more option that I really like the look of as well. So there's many different options for agent teams, right? You got setup profiles, you can run telegram group chats, you can have paperclip set up, and you can also have something called Hermes workspace. And Hermes workspace is like a mission control where you can have an agent team working together. So let me show you how that works. This is again, another open source project. So what you do over here is you're going to grab this information. You can just run another Hermes session. All right. So inside this session here, we can say, okay, set this up to run locally with Hermes. So this is going to be a mission control for me. And now we've got Hermes setting up paperclip, right? So this is setting up paperclip with Hermes. I'll show you the dashboard in a minute. And then we also have inside this tab, we have Hermes workspace being set up. And you see it asks for permissions before setting this up, but that's basically how you can do it. The final way for agent teams is you can use something called Multica. All right. So we're setting up paperclip here. We're setting up Hermes workspace over here. And then the final one is Multica, right? Multica is another way to manage agent teams and assign them tasks and get them working together. And so you might be asking as well, like, why would you have agent teams? Well, honestly, like if you just want them to work as like a sort of swarm of agents working together, then you would use something like paperclip or Hermes workspace or Multica, right? But for most people, you don't need to go that far. You can just stick to setting up Hermes normally, figuring out what your use cases are, giving it a good memory based on your specific context. And then from here, you can finally set up the agent teams. If you're like, oh, you know, I need more power. I need more agents working together. Or I just want to experiment and learn more about this stuff. Right. That's another option. So we got paperclip over here. We have this over here. Let me just check this is working. And then we have Multica to be set up here. Right. So you see how the agent can just take these open source projects and connect to them directly. Also, you can have multiple versions of the same agent working together as well. Right. Which is what we've got here. So we've got multiple agents working alongside each other, setting these all up. Right. As you can see. And then also, when you set this up, you want to make sure that you tell Hermes to connect with the setup as well. So if you're setting up paperclip or Multica or any sort of agent teams set up, make sure that you get Hermes specifically to connect to it. Otherwise, you won't be able to interact with Hermes inside these local models. So Hermes workspace should be working now. Let's just double check it. You can see here if you have any issues set up, then just give it the error and troubleshoot until you fix it. Now you can see it's setting up Docker for me. Right. So you can see how this is opening up Docker for Multica right here. And it's going to set that up for me as well. All right. So if we want to run paperclip with this, we've got it running here. Same for Multica. We can just turn that on and then we can connect to Hermes agent here, as you can see. All right. So Multica is now set up to and Hermes workspace is set up to. Right. So you've got Hermes workspace here. If we go back to paperclip, let's just make sure that's working. And then also Multica too. Let's just go to Multica and check that what's working here. So if we go to containers, Multica, we've got the local host there. That's still being set up, I think. So that will take a little while to run, but it's running the terminal commands. But in the meantime, we have Hermes workspace here. Right. And so if we have a look here, if you wanted teams to work together, you can go to the conductor or the operation section. Right. And inside those, that's how you can set these up directly. So that's it for Hermes teams. Let's see what questions we've got here. You are all right. Thank you very much, sir. Already testing Kimi K2.0 6 in Hermes and the difference is clear. Test memory of Honcho. I don't really use Honcho to be honest. Like I just find Obsidian is way easier to use with Omi. So I don't really use Honcho that much, but I'd be interested to know. I haven't heard that many people say how good it is with Honcho. So that's the only difference. And also like Obsidian is free. Alarma also have cloud models on the subscription. Yeah. So for example, like we're using Kimi K2.0 5 cloud with this. And Modric says, appreciate the live. Julian, thank you. Happy to help sir. Thank you very much. So just to recap so that you understand where we're at here. We've talked about the setup and how to use Hermes setup, right? And all the commands you can use. We've talked about the different use cases and how to make them personal to you using this prompt, right? We've talked about the memory and how to improve the memory so that it can be 10 times better because it understands you, understands your business, understands what you do day to day. And it helps you automate that. And we've also talked about how to set up agent teams, whether that's with profiles, telegram, paperclip, Hermes or workspace, right? Loads of different options for doing that. And then also you can see here, we've got Multica running locally. So we're putting our details here and then we can get set up again, like any sort of errors with this or any sort of problems, just ask Multica to get it fixed for you. But that's basically how you can do it. Another question we got was with Obsidian, does it use the entire vault or can you point at a specific section? So what you would do is you would go inside Hermes, right? And you just tell it like, listen, okay, focus on this folder or focus on this specific file. You don't use the whole of Obsidian and you tell it inside the instructions here. That's the easiest way to fix that. Now also with version B0.9, you can actually build something called a web dashboard, right? So you got Hermes workspace, which is pretty cool. But you also have this option here, right? Which is the Hermes dashboard, right? And so what this allows you to do is have a full mission control for your Hermes dashboard. You can see your recent sessions. You see what's happening. You can see your live sessions here. You got your analytics in terms of how many tokens you used, logs, cron jobs, right? So these are scheduled tasks that you can set up day to day. You can manage those directly. And you've also got your skills, right? So you can switch between all these different skills, see what's active, see what's not active, delete anything that you want to as well, re-enable stuff that you want to, right? So you can see for example here, we can switch that on and off. And then over here, we have a config, right? So we can change the models. We can see our agents, the terminal, the display, delegation. So delegation allows you to change models here, as you can see. And then you've also got your memory set up here too, right? Your compression, browser set up. You can add voice as well, text to speech, speech to text, et cetera, right? So there's some cool stuff right there. And then over here, you can manage all your API keys, right? And all the models that are set up with this as well. So this is quite nice too. So to do that, you just go inside Hermes terminal and you type in Hermes dashboard. And from there, you can get access to it. So Hermes is not even on version 1.0, right? It's on version 0.9. So the latest update of Hermes is still not even like the full version. So imagine when this fully launches as Hermes 1.0, how powerful that's going to be, how much better it's going to be. It's already super powerful and really easy to use, but it's going to get better, right? This is the worst that it's going to be. And already it's pretty powerful and easy to use. First time seeing Multicut also. Yeah, it's pretty cool. Pretty cool. All right. So we've talked about Asian teams. We've talked about memory. We've talked about use cases, and we've talked about how to build an automated thing with this. One thing that I want to show you as well, and also the dashboard too. And from there, you can manage your skills, et cetera, right? Plus scheduled tasks. So at this point, you might be going, okay, like for the scheduled tasks and for Hermes dashboard, what can you do inside there? What's useful? There's two big things that I would say are super useful, right? And so if we go back to Hermes dashboard here, there's two things. Number one is scheduled tasks, right? So scheduled tasks, for example, is like daily automations that run 24-7 without you autonomously. So, you know, that could be, for example, like researching news, posting content, managing or messaging your team with new updates, right? Stuff like that. So that's super useful. And then also you have skills. Now, skills are like individual experts that you can call upon anytime you need them, right? Now, these are like saved automations. So you can say, okay, do this, you know, and then I'll come back to you later. All right. So for example, you could say, okay, inside Hermes, you could say, okay, use your SEO skill to create content, publish it, index it, blah, blah, blah, right? And so that's how skills work. So scheduled tasks run 24-7 without you. Skills or saved automations you can call upon when you need them, right? So for example, publishing a new article. That's a great example. And so that's really cool too. And the way this opens up, as you can see right here, is it just like, it gives you a local host URL. So that's it for the Hermes dashboard. And then the final thing that I want to show you is the new update from Olama, one click setup. So if you're trying to set this up and you're like, oh, it's too difficult. I can't do it. No problem. All right. So what you can do here is you can use Olama and you would install Olama like this. It runs local and cloud models. So you paste that into your terminal. And then once you've done that, you can then use Olama, launch Hermes inside your terminal to run this directly, right? So you can run Olama launch Hermes and use free plus cloud models, plus set up Hermes in one click. So the benefits, you can use free models, cloud models, and you can set up Hermes in one click, right? So again, these are the benefits. People say like, why would you use Olama? This is why, right? Plus you can easily switch between models. There's also another option if you're struggling to set this stuff up, which is Max Hermes, right? So with Max Hermes, I've tested it. It's not as good as running it locally. It's a little bit buggy sometimes, but you can still do scheduled tasks. You can still use all the tools. It can create images and video straight out of the box for you as well with Hermes agent if you use Max Hermes. So that's another good option as well, right? It's another one click install. Easy to set up plus non-technical. So you've got two options for easy one click setups there too. Olama and Max Hermes. So let me show you how to set up Hermes with Olama quickly. So you just go inside the chat. You take this command, Olama launch Hermes, and then you select your model and then boom, shakalaka. You got Hermes ready to go right there. Right? So if anyone's non-technical and they're watching this, it's like, okay, just use Olama because you can install it in one click or use Max Hermes because you can install it in one click. And those are two great options right there. So just to recap, that is how to automate and build anything. So you want to set it up. Then you want to figure out, okay, what am I using it for? Everyone has different use cases. Here's some mine. If you're like, oh, I don't know what to do. Here's some examples of what we can do with Hermes based on what it knows about me. If you're wondering, okay, how do you get context and personalized information so you don't have like an AI agent that knows nothing about you, then you would use the Obsidian and Omi setup like I've talked about today. From there, you would set up agent teams. You want to get more power. You can use paperclip Hermes workspace, Multica, and set up multiple profiles, or you can have a telegram group chat. And then you can set up a Hermes dashboard with mission control over here using this command, Hermes dashboard inside the terminal. And you've got easy one click setups. If you're like, I don't know how to do any of this, right? And that's basically how to use Hermes, how to build or automate anything with it. And pretty much everything that I know about Hermes from, from testing it for, you know, hundreds of hours at this point, if you want to get all my best trainings and the full guide on this stuff, feel free to get it inside the internet. A.I. Profit Boardroom. This is my AI automation community. It helps you save time, grow and scale with AI automation. Inside the community, you can ask questions, get help and support whenever you need to. You can, me personally, you can jump on weekly coaching calls where you can ask questions about your setup, share your screen, and we go deep on agents like Hermes, OpenClaw, et cetera. You can meet people in your local city who are using Hermes and AI agents like this, as you can see right here. And then also inside the classroom, if you want to get my full guide and all the notes from today, I'm going to add that as a new page. Plus we have a two hour course on Hermes over here and that will come with the video tutorial from today as well. So you can get all the video notes from today inside the AI Profit Boardroom. Just go to the classroom, go to new and advanced daily tutorials, which is where I drop new stuff all the time. You can get my full six hour course on OpenClaw right there, full two hour course on Hermes, all my different AI agent workflows, as you can see. And this is all inside the AI Profit Boardroom. Link in the comment description or just go to the AI Profit Boardroom.com. Thanks for watching. I'll see you on the next one.