Hermes Agent for Beginners...Safe Set Up in Under 17 Minutes

I've been working with Hermes. This is an autonomous persistent agent that runs 24/7. And I gave it access to everything, my Mac, my Apple account, my Google and Gmail account. I even gave it its own credit card with a $10,000 limit and sent it on a mission to build a side business that is guaranteed to 100x my revenue. Sorry, everyone. I couldn't help myself. No, that's not what we do here. This is not a AI hype bro channel and if alarm bells are starting to ring, that's good news. Those are your survival instincts kicking in. Personally, I am very cautious when it comes to tools like this. And if you are as well and want to learn how to get started with Hermes but in a non-overwhelming AI hype bro kind of way, then this video is for you. On the real, I have been using Hermes and it's really good and crazy powerful. So, what we're going to do here is I'm going to show you how to get started safely, where to start at, and then also show some really good use cases that's helpful for us web designers and web creators. Lots of videos will show you to install Hermes right on your Mac. Just go over to this curl command, go into your terminal, paste it in, and install it. And this is for beginners. That is crazy. We need to be very careful when we are getting started using a tool like this and the kind of access that we give it. And this is why we're going to start off using Cloudways managed server. Now, this is a dedicated server for either OpenClaw or Hermes. And the thing about using this kind of a server is that it is built secure from the ground up. In fact, if you go through the landing page, you're going to see the main focus is on security. Hosting it locally on your machine it means it does get full access to your Mac. Now, go ahead and go buy your own Mac Studio or Mac Mini, but we're just looking at getting started setting this up and beginning to learn how to use this kind of an agent. Now, here are the prices. Doesn't cost much and I do want to be 1,000% transparent. I did partner with Cloudways on this video, but it's only because I have used Cloudways for many, many years running my own web design agency, and I only partner with platforms that I actually use. I do have a coupon code if you want to even get a better discount. So, all that will be inside of the description, but let's go ahead and get started. You'll see how all of this works. Once you go through the process of getting set up, you're going to be inside of your back end with Cloudways. You're going to go over here to AI agents, and then just add in your first agent. We're going to select on Hermes. Go ahead and give it a name. I'm going to call this just my agent. We're going to keep it simple, and choose your location. Since I'm in Asia, I'll choose Singapore, and go to continue. You could choose the size that you want and the RAM. This does matter because one of the powerful features behind Hermes is the memory, and the memory needs RAM. Now, you could go ahead and start off small, and you could scale this up as you need. So, I'm just going to leave it on the scout option just to get started. For your LLM provider, you got several different options. You could go ahead and put in your API key and get started that way. Right now though, what I'm going to do inside of this video is skip it because I already have a $20 a month OpenAI or ChatGPT plan. We're going to use that without having to use the API and pay by token usage. Unfortunately, Anthropic or Claude, they don't allow this anymore. So, we're going to stick with OpenAI. And if you have a basic plan, just a $20 a month plan, then this is going to be your cheapest option to get started. We're going to go to deploy agent, and it's going to take a few minutes. And now it is ready to go. Let's go ahead and click on it and take a look at what we got. And here we are. Now, real quick down at the bottom, you're going to see we have our connections for our channels, and this part is really important because this is one of the powerful features and functions inside of something like Hermes that we could connect it and actually work with any device wherever we're at. Personally, I like to use Slack. I know many people use Telegram. I use Slack because this is where my work lives and that means if I'm out on the go, something comes up, I could connect with my agent directly through my phone and continue working with my agent wherever I am at. We're going to come back to this in a moment. First, let's go ahead and launch the app and take a look at it. We are going to need to get our password here. Let's go ahead and copy that and select on open agent. It is ready to go. Just add in your password and now we are here. It does give you a wizard to get started. I'm just going to click on through this and we're not going to use any API keys yet. I'm going to show how to get started with open AI first just to get started with this, but let's go ahead and open up Hermes and see what we are working with. Now, if you're using like the clock code app or code X app, even chat GPT app, a lot of this is going to look familiar. We just got our chat. We got our option to choose our models. We even got our efforts. So, a lot of this is going to be very familiar. But, there's actually a whole lot more here. First off, we got our tasks. We're going to be coming back to this when we set up our crons. Now, think of the crons as like setting up a uptime monitor for a website. We could go ahead and set up a cron saying, "Hey, check this website every single 2 minutes letting us know if it ever goes down." We're actually going to do that inside of this video. We have our Kanban. We have our skills. So, one of the powers inside of Hermes is the ability to create skills. Hermes could create its own skills and there is a dedicated skill hub as well. I'll leave a link to this and all the resources that you're seeing. This is one of those things that just makes it really powerful. The next thing that makes Hermes powerful is its memory. It's the ability to remember what it is doing. It keeps everything persistent. Now, we have things I could go ahead and create notes and create different user profiles. It also has something called agent soul, which I think is really weird to call this since I believe only us humans have souls. Maybe uh should be changed to agent personality because of that is what it is. Uh we also have our project context. Something we definitely want to go through, but this and our skills, these are where the powers really come in and what makes a tool like this so powerful. Now, we also have our spaces and our profiles. We could create different workspaces, our to-do's, insights, and logs. As you can see, there's a lot. I mean, I could do a 5-hour course on this right now, most definitely, because there's so much, but that's overwhelming. That's not the point of this video. It's to get started in a very safe way. Let's go over to our settings because this is where I always like to start. Now, we got our conversation. We also got our appearance, and I'm going to switch it up because I do want to have something looking kind of cool. Do you know what? I'm going to change this to light. And this looks cool, but I'm kind of a dark mode guy, so I'm going to leave it on dark mode, and I'm going to put my font size a little bit larger. My eyes aren't what they used to be. So, go through the settings and your preferences as well. Now, our providers, this is where we are going to want to set things up next cuz we need to have a provider. Right now, it can't do anything. Hello, and there's no LLM provider here. So, let's connect our ChatGPT subscription. First, we got to go over to our ChatGPT, and there is a setting you need to turn off. So, we're going to go over to settings, and then over here in the security and login, go all the way down. We want to make sure we enable device code authorization for Codex. We need to go back to our Cloudways dashboard and get our SSH credentials. So, what I'm going to do is I'm just going to screenshot this cuz I'm going to take it into Claude and then I just ask Claude to walk me through connecting OpenAI Codex. We're going to need to open up terminal for this so we could SSH our way in and this is connecting to the server from our terminal. We are not installing Hermes or anything into our Mac that is going to be unsafe. So, let's just go ahead, open up terminal, paste this in and now it is going to ask us for a password. You're going to see in our SSH credentials, our password. Let's go ahead and copy that, paste that here. You're not going to see it. And the first time you do this it might ask you to say yes or no, just go ahead and type in yes if it asked you that. Now, we're going to go to the auth command. Let's go ahead and copy that, bring it back over and we are going to paste this in because now we are actually connected into our server. So, let's run this command and now it is going to give us a URL that we need to go to but it also gives us this code. This is the code that is going to authenticate the app. Go back over to my browser, choose my account and now we need to give it the code. I am going to copy this code and then paste this in and now my server is signed into Codex. Let's go back over to our Hermes agent and I am going to go to my settings, to my providers and you can see now it is already connected, it is green. Now, let's go back over to our chat and just put in hello, make sure we are connected. Okay, it's still not showing connected. Let's make sure we have the right provider. I am going to go to ChatGPT 5.5. I'm going to make sure I am inside of the Codex and let's just go ahead and refresh this and now we can see it is working. It is starting to process. We are now connected. Now there's one more thing we want to do before we start building out some automations or building out some tools inside of here and that is going back over to our backend dashboard and we want to connect one of our messaging providers. I'm going to connect slack. So the reason why we set this up first with our chat GPT is because now you could go in here and ask walk me through connecting to slack API and you don't have to go back over to your cloud or your open AI chat GPT apps. You could do everything right here and depending on how often you've done this, how experienced you are, this could take anywhere between like 5 minutes to 30 minutes. So just walk through the process like you would inside of cloud. I would walk you through that process but that's a dedicated a video in its own. Just use Hermes to walk you through it. But here is the awesome part here. Now this is the first thing that definitely you want to get started with. Here is my slack. Now I got a dedicated channel right here for my Hermes agent. I've already invited it to it so I could just put in hello at and then I can search for my Hermes agent and let's go ahead and see. Look at that. It is already working. Now we could connect together here and we could actually talk. Let's say I am working on something. Let's say we're working on a client project and we just need some information something looked at. I could ask my agent now to go do a task or fetch me some information or do something that I have already set up inside of Hermes. And if we go back over to our web app and we do have it right here. Jeffrey says, "Hello." And that means we could continue our chat here. We are connected, ready to go. So, what does all this mean? What can we do? Well, let's start with something actually very simple. Let's say we built a site for our client, and the site like what I have here. This is just something I vibe coded really quickly. Not like a real site. I just want to use it as an example, but this is a static HTML CSS site. Inside of WordPress, we got tools like ManageWP or WP Umbrella that could do the uptime monitoring. What if we need to do our own uptime monitoring on custom-built sites? I'm going to go back over to my agent, and I'm going to go inside of a new chat, and I can give it a simple prompt to set up uptime monitoring to the website. Create a cron that checks the uptime every 2 minutes, and if the site ever goes down, to send an alert message to our Slack. And after it's done, let's go ahead and run a practice alert. Let's go ahead and set that up. Now, right now it is running on ChatGPT 5.6, so this is going to be the highest effort. We don't even need to use this right now. I probably think 5.5 might be good enough. Some tasks, you're just going to want the lowest model, so this way it isn't using too many resources. But now, check this out. It does not take long. It is building everything out. Okay, and it's asking us which Slack destination. So, let me go ahead and put in the URL to the chat that I want it to go to. Let's send that. And now it is fully built. If we go back and look, it took 3 minutes, and the whole process, it went through, and it did its own fixing. It did its own thinking. It figured everything out and went through step by step. Now, setting up something like this inside of like an agent, and building an AI agent in there to just do something simple like this, if you've done it before, you know it could be a little bit complex and takes a little bit of time. This took 3 minutes and check this out. Inside of Slack, it did run the alerts. So, now we have the same thing set up like I would have in my WP umbrella for a WordPress website on a custom-built website. And this is where we're going to find it. We're going to go over here to our task and we're going to see our cron set up over here. This was built all with the script. There's no prompting needed. This is not going to take any tokens because it's all using a Python script and it built it out in just 3 minutes. We could go ahead and create more cron jobs ourselves and manually do it here. That's all good, but we could go back over to our chat and just create a new one. This Okay, I'm going to do something really silly right now. First off, I am going to do something I would not normally do, but just to show you how easy this is to set up. I'm going to call this one say hello. Just very very silly. Okay, I'm going to keep this on private for right now. I don't want anybody else seeing this and I am going to invite my Hermes agent. By the way, this is what I set up inside of Slack when I created my Slack app. And now I am going to just give it a very simple prompt and this one's going to create a new task every hour that just tells me hello and shares the weather inside of my city. very silly and I just gave it a URL to a new Slack channel that I started right here. So, it's going to set it up. I just want to use this as an example and it looks like it finished this in 14 seconds. No way. Let's go back and see. Did it do it in 14 seconds? I believe it did. You see, it created a prompt here. We could go in and we update the prompt and we could work with this cron task, this automation. Let's go ahead and close it. Let's run it and we have an error. Let's go back over and just let it know, "Hey, there's an error. Let's figure this out. Keep running tasks until it is fixed and it is working." This could give a good idea of where to get started. When I get started with a new tool like this, I first like to start small and look for the low-hanging fruit, the uptime monitoring. Imagine creating automated backups or automated comments to GitHub. A lot of different situations where I could see this being really useful. Even being able to have a dedicated agent getting to know your client and the project and the way you work together. Not as in you're creating an agent to give to your client, but uh agent that just has a memory and knows the way your client likes to work, how you interact about their brand and their goals. And that is your dedicated agent that is your collaborator to work with your client. Maybe your client is very active. They're getting a lot of reviews on Google, Trustpilot, Facebook, and so on. You could always offer reputation management services where you're creating an agent that runs on the cron, that checks and looks for new reviews or is triggered whenever it gets a new review. It could respond to the positive ones. It could flag the negative ones. It can make sure the mentions of the brand are all being flagged, they all are being looked at, taken care of, and making sure nothing bad is happening with the brand. And if there is a negative review, the brand is able to respond to it. Things like this could come in handy for us web creators. But what I found is when I started here inside of Hermes and get started with the first smaller task, the simple things, things that might seem very simple. Well, the more I began using it, the more opportunities I found like, "Hey, this could be really useful for this, for that. I could be taking a shower or driving somewhere and new ideas will pop up on how I could simplify my own workflows. And with that said, we can see it fixed everything. It took a couple minutes, 2 and 1/2 minutes, but if I look at my Slack here, here we go. Hello, says the weather. Okay, we got some rain today. And now we got a very simple cron started. This is just a way to begin seeing how we could use this inside of our workflows and being able to use something like this inside of Cloudways, that's very simple to set up. Just a couple things we need to do and doing it all on something safe. So, yeah, check out the links, the coupon code, all that stuff, all the information you need, they're all inside of the description of this video. I definitely would love to hear your thoughts on it. Again, we're just getting started. If you want to see more, which I would be so happy to do because to me, this is this is actually powerful and exciting. So, if you would like to see more, let me know. I'll be happy to create more and go into everything deeper with something like Hermes and how I'm using it and how this could be useful for us web creators. That's it for this video. Thank you for watching. I'll see you in the next one.