Hermes Agent: Build & Automate Anything for Free (Memory, Paperclip Teams, Scheduled Tasks + Free Models)The video explains how to use the open-source Hermes Agent to build and automate workflows for free, including setting it up from GitHub and running free models via News Research Portal (e.g., DeepSeek V4 Flash and Step 3.5 Flash) or other options like OpenRouter’s free Owl Alpha and local models with Ollama. It demonstrates installing and using Paperclip to organize a team of autonomous agents, assigning goals and issues, and checking the inbox for updates. For stronger “never-forget” memory, it recommends combining Omi (screen/mic note-taking) with Obsidian, exporting daily notes, and using Claude or Hermes to reorganize the vault into a structured system (e.g., PARA) with a knowledge graph. It also covers making agents proactive with scheduled tasks, setting up Claude MCP with Hermes, troubleshooting Hermes Workspace sync issues, deployment/privacy options (separate user, VPS, or cloud one-click deploy), and ways to offer Hermes setups as a service focused on lead and customer outcomes.00:00 Free AI Employee Intro01:08 Paperclip Team Setup04:35 Selling Hermes Services06:51 Run Hermes Free Models08:03 Career Focus Timeline10:19 Second Brain Memory Stack14:34 Scheduled Tasks Autonomy17:57 Claude MCP Control19:16 Token Saving Options21:23 Safe Deployment Choices22:41 Overwhelm Focus System24:24 Fix Hermes Workspace Sync26:55 Local Models For Websites28:20 LangGraph Integration29:03 Wrap Up And Community
Full transcript
Hermes agent can build and automate anything for free and today I'm going to show you exactly how to do it. By the end of this video you're going to have a 24-7 AI employee that works whilst you sleep, one that remembers everything about you and your business, one that stops waiting for you to tell it what to do and just goes off and does it on its own. I'm also going to show you how to get Hermes AI running completely for free, how to give it a memory so powerful it never forgets a thing, how to turn it into a fully autonomous agent that builds, creates and automates for you all day and all night and I'm going to show you how to take all of this and use it to get more leads, more customers and more clients. There's one thing I'm going to show you later in this video that most Hermes users have never set up and once you do your agent stops sitting around waiting for you, it just works on its own autonomously every single day.
So stick with me to the end because that one thing alone is going to change how you use Hermes agent forever. Let's get into it. Hermes agent, how to build anything. So today we're going to run through some of the latest questions we've had on Hermes agent, answer them, show you exactly how to build and automate anything with Hermes agents.
So let's kick it off. Ted here says installing Paperclip and Hermes agents. Paperclip is a way to manage like a team with Hermes agent and then you can create like a full autonomous team of AI agents. I'll show you that in a second.
It's a free open source project. Now you can see here that Ted has said, you know, imagine being set up, being interested in the Paperclip setup, especially for content creation, but not being technical at all. What would you do here and how would you get started? So this is a great question.
Here's how I would do it. I mean, number one, you have to make sure you've got Hermes installed. If you don't have to set up Hermes agent, you just go to the quick start instructions on the GitHub, right? Copy and paste them into your terminal.
You're good to go. This is terminal over here. Now for setting up Paperclip, you can basically, this is how you can do it. So if you're not technical at all, all you need to do is you copy this, paste it into Hermes and then you copy this GitHub as you can see right here and paste it into Hermes.
So there's two separate GitHubs for Paperclip. One for the Hermes connector and then one for Paperclip itself, both open source. So you can paste those into Hermes agent like so and just set and say set these up, right? And basically what that would do is it will take the information from GitHub on both projects and then go off and install it.
Now once you've done that, then with Paperclip, you can basically organize your team, right? You can organize your company. So the way this runs is once you've logged into Paperclip and you can see it's setting that up over here. Once you've set this up, you can actually give Paperclip a team, a mission and you just set up agents from there, right?
So it's pretty simple to set up and then once you log in, you define the goal, you set up the team and this is an example of what the dashboard looks like by the way. So this is Paperclip itself. Hermes has just set it up and launched it for me and essentially we set up a goal here. So we'd say, okay, like here's what I want you to do.
Here's my goal, grow the airport for the board and community, something like that. And then that's connected to Hermes now, right? And yeah, that's basically it. So just to recap on the steps, you take both of the GitHubs for Paperclip, you paste them into Hermes agent, you ask Hermes to install it, you give your agents a mission, which you can do inside Paperclip over here in the goal section.
You just set up a new goal like this. Here's an example of what a goal could look like. And then from here, you set up a new issue if you want to assign tasks to a specific agent. So we can for example, get Hermes engineer or Hermes test or the CEO to start working on the task.
And you also want to check your inbox as well regularly. And we actually have a training on this inside the AI Profit Board. And if you go to the new daily updates section here, and then go to Paperclip, you can see we've got full video training on it as well, right? So if you want to go really deep on that with like 100 prompts, a 30 day roadmap, covers all the limited beliefs and stuff like that, then you can get that right here, right?
But that's basically how you can use Hermes agent with Paperclip. And yeah, it's pretty cool. Pretty powerful stuff. Next question.
By the way, if you want me to ask, answer questions like this for you in a video tutorial, join the AI Profit Boardroom. And then every day inside the community, I answer questions like this in a sort of masterclass format where I cover all the questions everyone has, because I know if people post them here, there's probably, you know, 1000 or 10,000 people out there asking the same thing. Let's see what we got here. So Bambi has actually been setting up autonomous agents as swarms.
This is pretty cool. And here's an example of the UI as well. It looks super nice. This is a good question as well.
So Sean says, like, you know, when it comes to actually offering, for example, Hermes setups and Hermes installations, as a service, where would you start here? Right? This is really cool. So we have a lot of, you know, clients who come to us and they ask us to set up AI agents for them.
Here's what we look at. So I mean, the best thing you can do, number one, is speak to your clients, right? And ask them what they want specifically. So if you're, for example, focused on a specific niche or a certain industry, etc, speak to your clients, figure out what they want, what problems they have.
And now also figure out from there, what specifically the problems are that you can solve with Hermes. Some examples of like what we typically set up, they include, for example, like AI avatars, email outreach tools, anything related to lead generation. And what I've also seen when it comes to setting up agentic setups is like clients want anything that's focused on generating more leads and more customers. What they don't really care about is like the automation themselves.
They care about the end result, right? So for example, how many customers is going to bring in? How many leads is this going to bring in? Now, there's a few ways that you could sell this.
So for example, you've got Hermes one-click setups for installations, so you can set up everything for them. You can also offer like one-to-one consulting calls about Hermes agent and setting up, you know, you could, for example, set up an hourly process for that. And then also there's ongoing maintenance from there. So for example, if you set up Hermes agent for a client, well, then you could also offer a retainer monthly where clients hire you to maintain and set up new agentic workflows.
So I've talked about some examples of what we've set up. So AI avatars, email content, outreach tools, AI SEO is another one as well, which is helping clients rank on Google and inside AI search engines. Best thing you can do is speak to your clients to just figure out what they want specifically. When it comes to agentic setups, clients usually want something that generates more leads and customers.
And there's a few ways you can do this. You know, you've got the one-click setups, you've got one-to-one consulting calls, you could, for example, offer an hourly call rate, and then also you've got options where you could charge for ongoing maintenance as well. So those are some options. Now, Robert was asking, like, you know, do you need a subscription to create, for example, like a custom GPT or an agentic workflow?
It's quite a good question, right? So there's a lot of people out there who want to use AI, want to learn AI, but one of the biggest issues is that, for example, they don't have the resources to do it, right? So what's actually pretty cool about Hermes Agent specifically is that, number one, the project itself is open source. That means it's free and you can start using it today, right?
And then also, you can get free APIs for this. So, for example, Hermes just made an announcement today that Deep Seek v4 Flash will be free with Hermes Agent. You can see an example of that right here, right? So Deep Seek v4 Flash is now free via News Portal, thanks to Navita Labs.
So this is free to use. So if you want to set that up inside Hermes, once you've installed it, following the GitHub, you just go to Hermes Setup, like so, select News Portal, and then from there, you select the free model that you want, right? Now, it's going to give you the options here. So you can see, for example, we can use Step 3.5 Flash for free, and we can use Deep Seek for free.
And that way, you can still create workflows, but you don't need a paid account. This is a really good question here, right? So Arpiat says, like, what was the one thing you focused on at the beginning of your career? And for me, it went through stages, right?
So, for example, it's pretty interesting. So I started off with SEO, right? I learned the skill of, like, ranking websites. And bear in mind, like, every skill you learn year by year improves and compounds over time, right?
So it's the same, for example, with agents, right? It's like, I learned OpenCore at the start of the year. And then from there, I learned, for example, something like Hermes Agent. And the more you learn, and the more you stack your skills, the more powerful you become, right, as a person.
So I'll give you an example, like, as a timeline here. So 2017, I was really, like, focused on SEO. 2018, it was like building a, becoming a freelancer, I would say. 2019 was the year I started posting on YouTube, was focused on that.
2020, started building out my agency. In 2021, really is about, like, hiring people and learning how to manage people. In 2022, I actually didn't focus at all. It was really bad.
Actually, this was a year where I would work, like, one hour a day, right? So it wasn't the most productive year. It was super chill. It was good for learning how to delegate everything, but wouldn't recommend it.
I mean, it's good to have one year like that, just to see, you know, how things are, and if you actually like working or not. For me, I choose working over anything else. It's my favorite hobby. 2023, I started learning ChatGPT.
2024, I started really learning how to do YouTube videos consistently every single day, right? So I started posting daily tutorials, sometimes twice a day, etc. And that helped me a lot as a skill to learn. 2025, it was really the year of learning AI coding, you know?
So, for example, stuff like Bolt, Lovable, Virtual Studio Code, Quilt Code, etc. I was learning all of those skills. last year. And then this year was a year of learning open core and also Hermes agent.
Those are the two things and also learning how to grow a school community properly. So you know if you want like a timeline in terms of what I focused on year by year those were more. I had a focus pretty much one single focus for every year as I was growing. This is a really good question by Wayne.
So Wayne is asking he says hello everyone looking into every avenue possible to make my Hermes agent the best it can be. My main focus is memory right. I have obsidian and super memory installed but it still has lapses once in a while. Any suggestions on how to make the memory better?
So for me personally what I use is a combination of OMI and obsidian and then I connect that local vault to Hermes agent to open core into all my other AI agents. So you can see for example here like inside OMI it lists out my tasks my goals and it writes notes on me all day every day right and then I can export those notes to obsidian like so right and this is really good because what it's doing all day essentially is recording my microphone my screen writing out the tasks taking notes on me in terms of what I do daily and then from there I can export that to obsidian and when I look at your setup here with obsidian and super memory I would say the missing link here is OMI because it records your screen it records your microphone understands what you're working on all day it takes automatic notes based on that and then you can export those memories daily to obsidian. So if we go back to obsidian for example here what it's actually done is it's created like a pretty awesome memory that's super organized as you can see right here so as an index a daily template it takes notes on all my projects it's even got notes on Hermes and my series on that SEO size etc as you can see and it's super organized right and then we also have this graph knowledge view as you can see right here where everything is linked together and I can see exactly what links together how links together etc and so it's all organized it's all set up it's even got like a readme in terms of like how we organize this step by step the main structure of it the home and the readme markdown files rules for organizing it so AI agents know how to take notes where things should go and the style right and so we basically have a full setup for a second brain and this is the best memory system that I've seen so far right and so it pulls in all the best information it has resources about me my writing voice even my best playbooks and archive of me as well typical conventions and then it's just super organized and easy to use so the other thing that I would say here when you're organizing this whole memory system is a if you just rely on OMI and obsidian it's not going to be that organized what it's going to do is just create one memory section here right with a full index of all my memories as you can see right and in order I said just you know into like a raw export with 1200 memories right which is great but the problem with that is a it's all just one big long document so how do we get that working better well what we can do here is we can actually go up to Claude or Hermes right they both work the same way they can look at your obsidian vault and then organize it into a much better style right so if you want it to be super organized like I've shown today then what you want to do is make sure that you actually get Claude or Hermes agent to organize your vault as well and that way you can have a nice read me a nice home section everything links together the knowledge graph will be much more useful to G G rock says what app is that so this is obsidian so this is OMI and this is OMI here as you can see right that's what's taking notes from my screen in my microphone then you can use obsidian and obsidian is where we store all of our memories and have a nice knowledge graph like you can see and also a full vault of like all my memories organized by sections and projects and then finally we have Claude if you want to get it really organized it is something called the power system which is basically like project areas resources and archive right so it creates a nice little system right there if you want to go super in-depth on that we've actually got a full module called the infinite context engine over here right which gives you a breakdown of exactly how it works step-by-step a full guide 100 example prompts of how you can structure it and organize it but yeah that's really powerful so Karen says Hermes agent structure and setup right so first she says I'm super impressed by the community in June and gold is fast support recently joined and had all my questions answers so far to help her feel less blocked happy to help all right then she says I think I've been reading everyone's and I think I've been building agents upside down okay I think we all feel like that sometimes which is okay right so she says I always wanted to I always started with the orchestrator gave him a lot of context about me through voice notes and we're also sitting up sitting ball which is a great way to do they wanted him to handle stuff for me including creating a team of specialized agents so far it's always felt like I had to keep prompting him to do stuff all right but I'm thinking now I should create the specialized agents first does that make sense I would say I mean honestly I would say like you don't need to right let's run through this so when it comes to organizing your Hermes agent right for 99% of people they actually don't need multiple profiles because Hermes can handle many things on its own unless you're trying to build like something really complicated and crazy which I won't recommend if you're not sure about Asian structures first of all so for example if you look at this situation here Karen is saying like I feel like it's always me who has to keep prompting the agent to do things right so if you want your agent to do things without you autonomously then what you want to do is schedule in tasks because if you have scheduled tasks set up then your agent will do things without you let me show you an example that so if you're going to Hermes agent here and we type in Hermes if I'm like okay well I keep having to prompt and ask my agent instead of it being proactive to go off and for example create content for me well what I can do is schedule in a scheduled task instance inside Hermes agent and it will do autonomously without me every day right so I can say for example every day at 9 a.m. research the latest AI automation news right and if you always feel like it's you prompting your agent to do things then this is one way to run it right so you can set up scheduled tasks and you can see for example we've got an example right here right now if you want to cancel those tasks you can just say to Hermes agent cancel the tasks right now please cancel the task alright and then it was cancel any scheduled tasks now if you find yourself prompting the agent to proactively come up with ideas and new scheduled tasks that you don't have set up already and you're like I wish this agent was more autonomous what you can ask the agent to do is set up a scheduled task where it comes up with five ideas proactively to improve every day so I can ask Hermes agent here create a scheduled task that generates five proactive ideas each day to improve and help me automate more stuff right and you can see here so I'll set up what day each day should I do it right and so like that way you can set up the agent to run without you and it can be way more proactive and then you don't need to worry about creating like a content creator agent specialist you don't need to worry about orchestrators etc you can simplify it a lot so it solves two problems which is number one you simplify the process and number two your agent becomes way more proactive because it's actually got a daily scheduled in task where it has to be proactive to help you right now we've got a really good question from case in here which is like basically case and wants to set up Hermes with Claude MCP right to be able to control Hermes agent with Claude MCP so inside for example Claude as you can see right here if we go to our settings then we go to customize inside this section right and we can see that Hermes agent is set up here right now if you have a situation where you can't set up let's say for example we have a problem like this here's what I would do is I would open up a new Claude chat so open up a new session copy the information from the github like you can see right here then go back to Claude and just say like please can you help me set this up as an example of that you can actually see inside this chat that's exactly what we did so I said set up Hermes and Claude MCP and then he just went off and did it for me right I was just like just get it done so it worked pretty nicely and then bear in mind like it might take like 10 or 20 minutes to connect Hermes MCP with Claude but bear in mind once it's set up you get to go right and also it will give you details on how to set up the MCP how to install etc and then once it's done you can see an example here now this is a good question as well so Marlon says you know paperclip and token churn right like how to how to use agents without rinsing the tokens is there anything that you can keep using with paperclip in Hermes agent that doesn't use up too many tokens so here's a way that I would look at it you can actually get free API's inside Hermes agent like I was saying before so you got step 3.5 flash and you also have noose research now if you do use noose research portal inside Hermes sometimes even the free API's will be limited right so for example deep-sea fee for flash and step 3.5 flash so how do you get around that so what you can also do is you can go over to open router and there's a bunch of free API's inside here right the main one that I would look at let's just make sure it's still available is our alpha right our alpha is a free API that you can set up and use with Hermes and you can see it's been the top app that's been used with this Hermes agent so I'd recommend that bear in mind because these APIs are free it often means that any prompts and completions you use will be logged by the provider to improve the model so you have to be aware that like you know any anything private I wouldn't run through these free APIs but that's an option and also you can run free local models as well right so for example you can actually set up free models with a llama if you have a really good setup for me I personally don't read even on a Mac Studio I don't really use local models because it can be quite slow right but you know something like Gemma for that's pretty lightweight for almost any setup it's designed for mobile devices it means that the outputs of your agent won't be as good but if you want a free model where you don't have to worry about tokens that's another good option too and the cool thing about a llama specifically is a it has a one-click setup with Gemma for and any other model on their platform so if you want to run for example like coin coin 3.6 with Hermes you install it like this once you've downloaded a llama and then you copy this command over here and run that through a llama and then you can run Hermes agent for free so that's a pretty good option too this is a great question so one question we've got here is like if you wanted to put Hermes agent on an everyday computer could you create a new user absolutely that's that's one way to set up right so if you're if you're thinking about running these models but you like you know what I don't want to give it like access to my local computer then you can set up Hermes agent on a separate user profile on your Mac right and that way it doesn't have access to anything else so that's a really good way to do it and you just follow the instructions from the github here right which you can see you can set up a new profile the other option that a lot of people use is they use something like Hostinger and it will install it on a VPS which is a virtual private server so that's option number two and then the final option is that you can use something like a one-click install one-click setup so for example a lot of people use like max Hermes and what you do inside max Hermes is the you can set up in the cloud with a one-click deploy Hermes agent in the cloud using minimax right and so that's another way because it doesn't have access to any of your data and if your connections doesn't have access to your computer or computer use etc and so you can have a one-click deploy in the cloud for that so three options there the one is a separate user profile number two is use a VPS on Hostinger and number three is use something like max Hermes and you just deploy in the cloud all right we've got a great question from Michael here which is like you know Michael just joined the AI profit boredom community and he was saying like you know right now he's focused on a few different things but he's also feeling like slightly overwhelmed and this is something that a lot of people feel when they're getting into AI because you know there's so many different options there's so many different agents there's so many different tutorials out there what do you do and where do you focus right and for me what I would do I would recommend that you look at what you do day-to-day and then how can you automate that what's one thing you could automate and just focus on that once per week right you don't want to overwhelm yourself you want to enjoy it enjoy the journey enjoy learning enjoy the process right and so the way that I look at this is I try and focus on one thing every single week block out all the other noise and then go from there so for example for us like at one point you know managing our emails inside my agency was difficult and so what we did instead is we automated that with Claude and Claude writes a draft and then we can quickly click approve and write another example would be looking at you know what I'm spending my time on and I was spending a lot of time before on content creation so I create that AI avatar for myself and that helps save a lot of time too so whatever your bottleneck is in terms of time that's where you want to focus on now we actually have inside the community a overwhelm protocol right it's a step-by- step process which you can see right here and basically it's designed to help you never worry about overwhelm again right and so it's a six-minute training shows you exactly how to focus how I stay focused and never feel overwhelmed and if you follow that process inside here you never feel overwhelmed again because you just 100% you know in the zone right you don't have to worry anything so that's why we set this up for members here now also some people struggle with setting up like extensions inside Hermes agent let me show an example so a great example of that is Hermes workspace right so this is a free extension for Hermes agent it allows you to manage your agent in cloud in a beautiful user interface but the biggest issue is that as great as it is right it sometimes doesn't sync properly sometimes doesn't work properly etc right so here's what I do instead what I actually do is I will take the github information like so from Hermes workspace for example and I would just say set this up right paste it in and say set this up now if you've already installed it but it's not syncing properly then there's two options I would say I would go into the terminal with Hermes agent I would just say it's Hermes workspace working right and just make sure if I've installed it is it synced properly now what you can also do in the background is you can go to something like Claude as well Claude can help you fix issues like this and you can say I've set this up but is it working with Hermes agent locally please check and test it for me right and so that's two different options you can get Claude to help you but you can also get Hermes agent to help you as well and you can have them working in parallel at the same time if you want so let's just find if this is working here so you can see here it's connecting like so and then we have Hermes agent now if I have any problems I'm gonna take the problem that have inside Hermes workspace feed that into Hermes agent and just say hey can you help me set this up and you know and it might take a couple of minutes to load up sometimes Hermes workspaces work can be a bit buggy because it is just like a free open source project so you need to be aware of that too so just to recap Hermes workspace is a little bit buggy so just be aware of that paste in the details from github for workspace into Hermes agent and get it to fix it for you from there and also ask Claude if you're still struggling to help you fix it locally too and that way you've got two agents in parallel helping you fix and sync it and then if you have any issues whilst you're using Hermes workspace take that issue describe it inside Hermes agent or inside Claude along with the github details and ask it to fix it for you based on the issues and also get it to test it for you so that you can eventually get it set up for you that's the way that I would approach it right and you can see that we got it working nicely over here right so we've got the actually Claude said everything's working fine here's a status it's running on this gateway here it is blah blah blah right pretty simple and easy it is a little bit technical though bear in mind as well this is another question we got so a question from Neil is you know Neil said I installed Hermes desktop and I'm really impressed I'm trying to get it to build websites and I'm wondering which local model would be the best right the I mean for me personally for me personally I don't use local models that much because they're not that good especially on a Mac mini but if you have to pick one I would probably go I mean like if when it's not good enough for you then I probably go with something like Gemma for bear in mind these models are going to be lower quality now what you can also do is just use a free API in the cloud like our alpha or step 3.5 flash or deep seek flash v4 via news research portal and that way you know if you want the benefits of free API and you want to build websites with Hermes agent you can do it that way instead but again like I don't think local models are that great in general even when you have like a Mac mini or a Mac studio setup now if you wanted to build WordPress websites and Elementor Pro websites etc what you do is you would build them locally via like HTML and then you would give the for example if it's WordPress then you give the username and the password and the API token from WordPress directly see agent and the same for Elementor Pro I've never connected that to an API directly but that's how you could do it but with WordPress I've set up with many agents including open core and Hermes and it works really cleanly this is an interesting one so Mohammed was asking like how to connect Hermes agent with Landgraf now Landgraf is not something that I've used but I've heard of it so let's just double check this and I assume it's gonna be on on github yeah so you can see it right here so I would basically take that documentation like so and I'd paste it into Hermes agent and just say can you set this up for me now as I saw that Landgraf I think they have an MCP yeah you can see it right here right so Landgraf they have an MCP and so what you could do is you could actually take the information of the MCP go into Hermes agent and ask it to set up the MCP as well right and so that way you create like a bridge between Landgraf and between Hermes agents so they can communicate so that would be the other option as well so that is it for every question I've answered every single question about Hermes agent how to use it etc showing you tutorials showing you examples showing you how you can set everything up right and pretty awesome there's some awesome questions here if you want to get your question answered like this feel free to post inside the AI profit boardroom we have an amazing community it's super active lots of people posting lots of people sharing stuff which is awesome as well inside the calendar here you can jump on weekly coaching calls which is great so you can jump on weekly coaching calls get help and support share your screen in real time inside the classroom you access all of my new daily tutorial setups tutorials extra you can see for example we have trainings on like how to go from beginner to expert with AI automation here we have new daily tutorials in advance setups like so for example Claude MCP with Hermes paperclip of Hermes Hermes desktop etc if you need to know about Hermes agents and agents in general which is great and then also we have a training on how to get more clients all of my best automations are inside section so if you've ever wondered like how do I create AI avatar videos or how do I automate X and that sort of thing you can check that out all inside here you get all my best playbooks for that you can also learn how to get more clients through the agency course and how to rank number one with AI SEO over there plus any coaching calls we record and put them inside here in case you miss them and also we have a map where you can connect with local people in your city right so you can connect with people in your city who are using AI agents and you know just meet them which I think is awesome because it's quite hard to meet people who are interested in similar stuff like this so thanks so much for watching hope to see you in the next one and that is how to use Hermes agent link in the comments description or go to the aiprofitball.com to get access
More episodes