---
title: 'Best Second Brain Setup 2026: Obsidian + Hermes Agent'
source: 'https://youtube.com/watch?v=Q0HTefP9DFU'
video_id: 'Q0HTefP9DFU'
date: 2026-07-14
duration_sec: 0
---

# Best Second Brain Setup 2026: Obsidian + Hermes Agent

> Source: [Best Second Brain Setup 2026: Obsidian + Hermes Agent](https://youtube.com/watch?v=Q0HTefP9DFU)

## Summary

This video demonstrates how to integrate Obsidian with Hermes Agent to create a powerful 'second brain' that is both human-readable and AI-accessible. The presenter explains the concept of 'living files' and provides a step-by-step guide to setting up a VPS, installing Hermes Agent, and syncing an Obsidian vault across devices for maximum productivity.

### Key Points

- **Why Obsidian + Hermes Agent is the best second brain setup** [00:00] — Connecting Hermes Agent to Obsidian allows users to visually see what the agent is doing, customize skills by editing Obsidian notes, and compound knowledge without repeating tasks.
- **Living files vs dead files** [01:30] — Dead files cannot be accessed by AI agents; living files are markdown files that AI agents can use as skills, reference, memory, or context. Obsidian vaults become living files when synced with an AI agent.
- **Problems solved by this approach** [04:00] — Visualizing and customizing skills becomes easy; skills are kept up-to-date; agent work becomes visible and beneficial; AI outputs are saved as living files for future use.
- **Memory architecture: system prompt vs on-demand knowledge** [05:30] — Core memory (system prompt) is always loaded; notes and skills are loaded on demand when relevant, preventing context window bloat.
- **Setting up a VPS with Hostinger** [06:30] — Purchase a KVM 2 VPS from Hostinger (use code David for 10% off). SSH into the server using iTerm or terminal.
- **Using Pi Agent to manage the VPS** [09:00] — Pi Agent (running Opus 4.5) is used to install Hermes Agent on the VPS by pasting the quick install command from GitHub. The agent monitors and fixes errors automatically.
- **Configuring Hermes Agent** [12:00] — Set up OpenRouter API key and select a model (e.g., GLM 5.2 or Kimi K2.7). The agent can be managed via plain English commands.
- **Installing Obsidian and setting up sync** [14:30] — Download Obsidian, create an account, purchase Obsidian Sync ($5/month), and create a new vault. Sync the vault between MacBook and VPS using Obsidian Sync or Syncthing.
- **Connecting Hermes Agent to Obsidian vault** [18:00] — Use Pi Agent to install Obsidian headless on the VPS, log in, and sync the vault. Hermes Agent can now access and manage all files in the vault.
- **Using slash-go for autonomous tasks** [21:00] — The slash-go feature allows Hermes Agent to work towards a verifiable goal (e.g., scraping YouTube videos) until completion, saving hours of manual work.

### Conclusion

Integrating Obsidian with Hermes Agent on a VPS creates a powerful, AI-managed second brain that turns static notes into living files. This setup enables autonomous task execution and knowledge compounding, giving users a significant productivity advantage.

## Transcript

So, this is Obsidian, and it has all the info about my business, life, and goals. And now, I added Hermes Agent to it for the best second brain setup you could possibly have. And in this video, I'm going to show you exactly how I set this up, why I believe this is the best second brain setup possible in 2026, and several use cases that you can start implementing today to be more productive and more organized. So, first, let's talk about why this actually makes Hermes Agent a lot better. And the primary thing is that you, as the user, the human, you can see what the agent is doing. Normally, it works on files that you never look at, but connecting it to Obsidian and having all these visual graphs and these connectors, seeing the patterns, having a clear folder, clear markdown files, easy to edit, all of that makes you, the human, more in charge. It makes you a better orchestrator. Also, you can customize anything by just editing an Obsidian note, right? You don't have to go into the code, config files, terminal. If you just get an idea, you just edit it as a skill, if you have a lot of skills. For example, right now, I have 185 skills myself. Roughly half of them are pre-built into Hermes. The other half are my custom skills. But, if you want to edit something specific, Obsidian is probably one of the best interfaces to do it in. And your AI agent will use the new version of that file instantly. Also, your knowledge compounds. Every time you have to repeat, like, remember this, save this, all that stuff, it's wasted tokens, right? But, if you have it nicely organized inside of Obsidian in your second brain, then you can easily reference specific files. You can see what you already have stored. You don't have to keep asking the AI. You don't have to do the same research you already did. Plus, this also allows you to have the same files on multiple different devices. So, for example, here, uh I have these files on my MacBook, but with Obsidian Sync, which I'm going to show you how to set that up, I can have it on my phone and have it on a VPS where the Hermes Agent will be running. So, these files can be in multiple locations. So, even if someone steals your phone or your MacBook gets destroyed, your knowledge, your SOPs, your playbooks, your skills, that will remain. Now, perhaps the most important concept to understand here is the difference between dead files and living files. Most of the files that you have currently are dead files. What that means is that no AI agent can access them. And that's because they cannot work on them, they cannot use them as context, they cannot compare them to other files, they cannot do research on them. They're dead files. And that doesn't matter if they're on a disconnected hard drive or if they're sitting on your computer or in your Google Drive. If there is no easy way for a powerful AI agent to access them, those files are useless. So, in order for any file to be considered a living file, which by the way, that's like a term that I started using 2 months ago, and I guess it kind of caught on, it needs to be able to be used by any AI agent as a skill, reference, memory, context, part of a prompt, anything. It just needs to be usable by AI agent to do tasks more efficiently, to do work for you, to be more relevant and helpful for your life. If a file can do that, if it can be accessed by AI agent, it is a living file. So, in this video, I'm going to show you how you can take your files and using Obsidian and Hermes agent, turn them into living files so that AI agents can actually work for you. They can actually improve your life and so that you can have better context engineering than 99% of people. Living files are just markdown files, right? They can be used by any AI agent and all of them basically run on markdown files. So, you can take your Obsidian vault and load it up in open code, cloud code, code corrects, or any other AI agent and it can instantly see your organized second brain and you don't have to do any of the organizing yourself. This is why a lot of people got intimidated with this like previous wave of these Notion content creators, all this Obsidian stuff, is because it requires so much work yourself, right? You have to have perfect awareness of every single folder, of every single file, the entire structure. All of that is a thing of the past. Now, we can use a powerful AI agent such as Hermes agent to do all of that for you. So, you can focus on the thing that you do best, generating ideas, writing down your unique thoughts, and doing the things you actually like doing, and outsourcing everything else to Hermes agent. Also, there are several problems that this approach fixes. First of all, it was very difficult visualizing and customizing skills in Hermes. So, that's the first problem. Second, most people never look at their skills at all, which means they have skills that they haven't used in months. The skills that they are using are outdated and not really relevant. Huge issue. Next, the agents work as invisible, right? If it creates a file somewhere, you kind of need to either remember it or you probably forget that it is created. But when you use Obsidian like this, you can see your whole vault, all the files, all the folders, the entire structure, or if you prefer the graph view, you can look at the graph view and see how everything is interconnected on a semantic level. On that point, usually the work that AI does, most people don't get any benefit from it. When you run like a deep research query, if you don't save it into a living file, aka a markdown file that is easily accessible, usually on a VPS so that multiple AI agents can access it, and the AI agent can run 24/7, the AI work is useless. You're paying for cloud, you're paying for ChatGPT, but if you're not doing something with their outputs, it's useless. Like, what is the result of a deep research in ChatGPT that you ran 3 months ago? It's not working for you. But if you think this is just relevant to skills, you are missing the bigger point. With your context, the model becomes more useful to you, and your notes are your context. This is what makes Obsidian very, very powerful. It's perhaps the best way to organize your notes, your thoughts, your context in a way that's both human-readable and agent-readable. But the agent doesn't silently read all your entire vault into the memory because that would be very inefficient. It would cost a lot of tokens. It would bloat the context window, right? There are two things that make it really work. First, there's a small part, which is the system prompt, the small memory that it always loads, right? The most essential things, your name, how you prefer your answers, like I always put in make your answers concise, stuff like that, right? In Hermes, this is known as the soul.md file. The second one is most of your knowledge. This is going to be your skills, your notes. These are only loaded on demand when it's relevant, right? So, these are not going to be blowing the Codex window unless you're talking about that specific thing. So, the way to understand it is that the memory, the core memory, the system prompt is always on, and your notes and skills are only pulled in when relevant. Now, let's focus on setting up Hermes agent. The best way to do that is on a VPS. You literally give a full computer to the AI agent so it can do anything. And the easiest VPS setup out there by far is Hostinger. They even have a dedicated landing page just for Hermes agent, which I'm going to put in the first link below the video. So, here when you get here, just click on choose plan and select KVM 2. This is more than enough to run Hermes agent, Open Claw, Pi agent, Agent Zero, Codex CLI all on the same VPS. Here, you get redirected to your cart. Select the period. I recommend going with 24 months because you get the best value, and you set it up once and you have it forever. It's boosting your productivity every single day of your life. And personally, every month that goes by, I'm adding more and more VPSs into my life and business. So, the sooner you'll get started on this, the more advantage you'll have over everyone else. Now, here we don't need this, so uncheck this because we're going to be using a different AI provider. However, on the right, click on have a coupon code and type in code David to get additional 10% off. Hostinger was kind enough to sponsor this video, so if you want the best deal possible for your VPS, make sure to use code David. Most of these add-ons we can skip. If you plan on scraping, you can check this off. If you need an email for agent, you can check this off, but you can leave most of them off to be fair. For the operating server, you can select either Hermes agent or go for plain Ubuntu VPS. But the server location, just select something close to you. So, I'm in Poland right now, Katowice. So, Lithuania is very close, and that's good. So, then click on continue. Let's go to the checkout. If you don't have an account on Hostinger, just create one. It's super easy, takes 20 seconds, and then finish the checkout by adding your billing details and your credit card information. So, once you purchase the VPS, go to the left and click on VPS, select the one you just purchased, and you'll see a bunch of stats. So, this since this is a fresh server, you can see that we don't have any usage data yet. But, what we do have is the SSH access, which allows us to connect to it from my own computer, right? You don't need to understand how SSH works, none of that. It's just basically a way to manage your VPS through your MacBook. So, I'm going to copy that, and we need to open that inside of a terminal. So, I'm going to type in terminal, boom, paste that in. And then it's going to ask you for your root password, which if you didn't save that, you can easily change it here. And generate. Just make sure you actually save it because we're going to be needing that. Go back to the terminal and paste that in. And now we're connected to our virtual private server from our MacBook. So, I can do clear. I can do LS. You can see the directory. PWD, we're in root. And here we can do basically anything with the server. So, now we can manage our own VPS and do anything we need with it. Run automations on it. Run different AI agents on it. Chrome jobs. Whatever you want. It's basically a full dedicated computer running in the cloud that you can utilize for so many things. So, again, if you don't have your own VPS server, click the first link below the video and set up your first VPS on Hostinger. This is what I use for all of my VPSs. It's what everyone on my team uses. So, again, click the first link below the video and give it a shot. And if you want the best deal possible, use the code David for another 10% off. All right, so I'm going to go back to the terminal. And actually, there's a better way than using the default macOS terminal, and that's using iTerm. So, if you go to iTerm.com, this is a free application that is basically a better terminal for the iTerm era that has a lot of different cool ways to save time, nice keyboard shortcuts. It's easier to use. It's faster and more pleasant to use than the default macOS terminal. So, if you don't have iTerm, definitely recommend downloading it. Again, it's open source and completely free. So, here's what it looks like when you open iTerm. I'm just going to zoom in a bit and CD into root level. And what I'm going to do is I'm going to open Pi Agent. Boom. And again, I have videos on all of this, so feel free to watch them after this one, but you can use any other agent, Claude, Code, CodeX, Hermes, whatever. We're going to use Pi to manage the VPS, to set up Hermes, set up Obsidian, and do everything we need on the VPS. So, instead of me having to be a DevOps expert, having to, you know, SSH into it and do all this stuff, we're going to use an AI agent to do all of that for us. So, again, I'm going to copy the IP address for the root access. Go into C Max, and on the right-hand terminal, I'm going to paste that in. And you can do all of these steps in the default terminal if you prefer the default Mac OS terminal for some reason, feel free to use it. I'm just going to show you my own workflow, which I've optimized over the last couple of years, right? So, click on uh change if you forgot the root password, or if you have it saved, just copy it. And then, back to C Max, paste in the root password, and we should be SSH'd. There we go. We're SSH'd to the VPS. So, I'm going to say, "Check the right pane, C Max pane, in this C Max workspace, and let me know if you see the VPS uh stats we just SSH'd." So, I'm going to confirm with Pi Agent that it can see this and it can access this. And that's the beauty of C Max. This Pi can easily see this right-hand side. Both of these are terminals. And yes, it can see it. So, I'm going to say, "Great. Now, run clear in there." So, instead of me typing clear, I can have Pi run the clear. Obviously, this is a very simple example, but it can do a lot more. We can, for example, run CodeX CLI on there to install CodeX agent, which I think is the best for like coding and managing your servers, or we could just delegate the Hermes setup installation to Pi. And you can talk to Pi Agent in plain English, which I have running uh Opus 4.5 because, unfortunately, Fable was disabled. But, any decent model will work. The main idea here is that you talk to an AI agent in plain English, and the agent manages the VPS and run runs all the terminal commands on the VPS. So, let me show you how powerful that is. I'm going to go to GitHub and to the official Hermes agent repository, which is almost at 200,000 stars. So, huge congrats to the News Research team, the guys who created Hermes agent. Amazing accomplishment. I'm going to scroll down and go to the quick install section. Copy it. Boom. And then basically go back to CMAX say, "Now install Hermes agent root level on this VPS." Boom. And again, I'm going to go do anything. I'm not going to I'm going to lift my hands so you guys can see I'm not doing anything. The Pi agent is going to monitor. As you can see, it goes into sleep when the setup is going. So, on my right-hand side, this is installing all the dependencies, packages, everything needed for Hermes agent to work root level on the VPS. And if anything is missing, Pi is going to figure it out. So, the Pi agent is monitoring the status of the VPS. It's monitoring the Hermes agent setup, and I don't have to understand any of it. I'm not going to even pretend I understand. Like most most of these dependencies and terminal commands, I don't even know, okay? Even though I'm managing like 15 different VPSs, and I've tried all the AI agents myself, and I have thousands of hours with these AI agents, I'm not a Linux expert. Like I'm not a DevOps professional. I don't know all these things. But, I can talk to Pi in plain English, and I know the goal I want, right? I know I want to set up Hermes agent with Obsidian, and that's enough. As long as you can speak English, and you know the goal you want to achieve, if you have a powerful agent like Hermes, like Pi, like Agent Zero, you can easily talk to him in plain English, and he can run the commands. And as you can see, when it's waiting, it just goes to sleep. Sleep 20, which means it will wait 20 seconds to see if there's any change on the VPS. And I can literally just go chill. I can go outside for a walk. I can go have a coffee. I can launch another Codex Loop agent in the background. Whatever I need to do, Pi is going to take care of it. There we go. And let's look at Can you actually interact with this? It's going to be interesting. Yes, so it's going to be asking me questions. So, when it doesn't know, it just asks you, right? So, that it it set up everything. So, now we have Hermes Agent and root level. And we can just go with a quick setup. So, I'm going to speed run through this. Boom, quick setup, waiting for approval. So, we don't we don't want to do news portal. Actually, I'm going to kill that. Boom, Hermes setup provider inference provider. So, actually, I'm building my own provider. I'm still keeping it private. We'll see if I release it to public if there's enough interest. The whole idea is complete token controls. Having full transparency and of where your tokens are going. Just observability graphs, you know, being able to set limits per model per API key, complete controls. So, I'm building that for myself and my team. I might make it public, we'll see. Depends if there's interest in it. But for now, we can go with open router cuz it's kind of classic. So, just go to openrouter.ai, create an account if you don't have it, and then go to credits, charge up some money. You don't need $200, just do five or $10. Then go to API keys, click on new key. I'm going to name it subscribe. If you're watching this and if you want me to make more videos on Hermes Agent, make sure to subscribe. In YouTube Studio, over 60% of you are not subscribed. So, majority of you are not. And it's completely free again. So, please go below video, click the red subscribe button. It helps out a lot. I'm going to put some limit, just $20 on this key. Create, copy, go back to C Max. And again, keep your API keys private. Do not share them with anybody, okay? And again, in C Max, you can easily re- resize these terminals, which is just amazing. And again, you can zoom in each terminal individually. So, one you can have like completely zoomed out, the other one completely zoomed in. It's really great. So, if you're still using the default Mac OS terminal, guys, just switch to C Max. It's open source, it's free. It's a no-brainer. I'm going to paste in the open router API key. Then we need to select the model. Now, normally, I would go with open source for me fast, but we need to go open source, guys. And GLM 5.2 is amazing. Probably the best open source model right now from Z AI. So, go with GLM 5.2. It's way more cost-efficient than Opus as well, and it's like the closest model to the cutting edge of uh of the closed-source AI models, right? So, if you haven't tried the GLM 5.2, I highly recommend you do so. It is the first open-source model that doesn't feel like it's behind. When it comes to this, I'm going to keep the current setup, just enter through it. We're not going to do any um any messaging platforms right here. That's good. Local browser, that's fine. Choose a provider for image generation, we can skip that. Skip. Skip. Skip. Okay, we're done. So, now we can switch back to having Pi in charge. Say, I went through the setup. Try launching Hermes and see if it works. The beauty of this is that if any of you encounter error, talking to Pi can fix it, right? So, this is the thing. Like, you can use agents to set up other agents. All right, so it started Hermes agent, and it's probably going to send a test prompt. Now, if I had to guess, send a test prompt. There we go. Hermes agent is working, and it's fully set up on a VPS. The next step is connecting Obsidian so that it's synced between your MacBook, your phone, your Obsidian cloud vault, and the VPS, so that Hermes agent can fully manage your second brain. Now, I can say like, ask We can either interact with you directly or through Pi. Maybe it's better to do it directly now. List out all your skills. And by the way, if you want to switch to a different model, maybe like Kimi K2.7 code, you can just copy the model name, and then inside of Pi, you can say, now switch the default model for Hermes to be instead via Open Router, right? So, the Pi can change the Hermes config. You don't need to know where is the YAML file, where that is located on the VPS, how Hermes looks the internally, the architecture. You can just tell Pi, "Hey, I want to change the default model," and it will do that. And now it's debugging the VPS way faster than I would be able to type these terminal commands and way way faster than I could understand all these thousands of tokens, right? Like imagine if you had to read all of this as a human. It would take you minutes a long time to read all of this and to understand all of this. The model, it can absorb 10,000, 50,000, 100,000 tokens and just instantly understand what it means. So, you can't really compete with the model managing your own VPS. So, a lot of you need to kind of give up your you know, needs to be a Linux expert DevOps DevOps expert and realize that if you have the right setup like I have inside of tmux with the left side being Pi Agent, the right side being SSH'd into the VPS, setting up setting up the Hermes Agent, you're just going to outcompete the people that are doing everything manually. Having the right Agentic setup is really the biggest productivity cheat code in 2026. So, now if I do clear and type in Hermes, we should see the model being switched to Kimiko 2.7. There we go. Let's do hey, who are you? What model are you? As you can see, it's running Kimiko. We even have some Chinese characters right here. So, yeah. That's how easy it is it is to manage your Hermes Agent, update your Hermes Agent. Stop trying to do everything manually. Just sit down for 30 minutes, create a nice proper setup that lets you do your work more efficiently. And just like this tmux workflow makes it very easy to manage a VPS, I'm going to show you how to manage your entire second brain with Hermes and Obsidian. So, next, let's set up the Obsidian. All right, so the next step is going to obsidian.md to install Obsidian. So, click on get Obsidian for macOS or if you have Windows, Linux, whatever. So, Obsidian is basically the best way to manage your markdown files and your context for AI. So, again, download the app. Then, go to account and make sure to create an account. And that's because we're going to be getting the Obsidian Sync, which is a subscription to sync it across all your devices and the cloud. It's an encrypted way, so even the people at Obsidian cannot access your notes, but this allows us to have a synced state of all our skills, all our markdown files, notes, everything between your MacBook, your VPS, your phone, how many however many devices you need basically. So, there we go. I just created a new account and signed in. Next, we need to go into Obsidian on your MacBook. So, this is what it already looks like when you have something open. As you can see, I have a bunch of markdown files. I have the graph editor, but I'm going to show you everything from scratch, setting up a new account and a new vault. So, first go to settings, bottom left. Click on general. And then here in the middle you see account. Click on login. And login with the same account you just created. Boom. And there we go. Login. And we're logged into the account. Amazing. Next, we need to create a new vault. So, if you go to the top and click on file, open vault. As you can see, I have two existing vaults here, but I'm going to create a completely new one. Click on create. Vault name is going to be Hermes plus Obsidian. Just put in some vault name. It doesn't matter much. Location, select the location on your MacBook where you want this vault to be saved. Vault two, create, open. Boom. And create. This is going to open a fresh vault. As you can see, okay, close the other one. And this is what it looks like. Welcome. This is your new vault. Create something, create a link, blah blah blah. So, on the right you can create a new note. You can name this for example, business roles. And then you can list out, you know, all your people. Name. Name. Name. On the right you can see the graph and you know, all the connections in there. This is to create a link. So, you can for example, link to like welcome. Add link. You can link to like different things. You know, name welcome, welcome. All this stuff is going to show up on your graph. But again, the main idea here is that we don't have to manage any of this. Hermes agent is going to do all of that for us. It's going to organize the best possible structure for your context, for your notes, for your skill, for your entire second brain. So, the next step is connecting Obsidian to the VPS so that Hermes agent can manage it fully and completely. So, to sync your files between your MacBook and the VPS, there are two options. One of them is the official one from Obsidian, that's the Obsidian Sync, which is uh paid $4 a month, or you have Syncthing, which is open source and free, which requires a bit more setup, so it's up to you. I'm going to link both of them in the pack. Again, second link below the video. Everything I mentioned, all the skills, setup, terminal commands, literally everything mentioned in this video is going to be in the second link below the video. So, feel free to grab that. It's going to be completely free. I'm going to go with the option because I think most of you are going to go with that, and it's the proper way. It's unencrypted encrypted. It's the easiest setup. It does cost a couple of dollars, but if you want to go with the open source Syncthing free alternative that requires more setup, feel free to do that. The steps are roughly the same. So, first, I'm going to purchase the monthly $5 plan here. There we go. So, the plan is active, so now we can open Obsidian to start syncing notes. Literally click this button, and it'll We're going to select the new vault here, and at the bottom below my webcam, you should see it actually. Not this way. You see this sync? It's red. We should initialize it. So, click on Obsidian settings, sync, choose the remote vault, create new vault, going to name it remote 02, region automatic, encryption end-to-end encryption, and set a clear password for this remote vault. Boom. Make sure it's a very secure password. If you're not using a password manager, please sit down for 20 minutes and set up a proper password manager. Create. Boom. Remote vault. Connect. Continue. Merge. We're going to put the same password here, and click start syncing. And here we're currently connected to the remote 02 vault. So, the next step is actually going to be installing the Obsidian headless on the VPS. So, what I did is I copied the full setup, full prompt, and again, this is going to be be in the bundle, second link below the video. Just put in your email and I'm going to email you all of these resources completely for free, so make sure to grab that. And I just paste it into pie and says, "Run this setup, right?" And it killed the Docker container, which was incorrect. And now it's using Obsidian headless to do it. It's asking me for the email, so I'm going to type that in. Boom. It's going to ask me for the password. Is this the account password or the remote vault one? Answer in short. When you're not sure, just ask the agent. Don't be afraid to ask. Okay, it's the account password. So, I'm going to copy that. Boom, just going to paste that in. Enter. And we're logged in. Check now. We should be logged in. And literally the agent walks you through the steps. It helps you figure out any debug issue. Listing out the remote vaults. Okay, it found the remote vault. There it is. Remote 02. Amazing. This is This is so good, guys. Like, I highly recommend doing the C-Max setup by agent on the left or any other agent. Okay, now it wants the encryption password. And again, I'm pasting it directly into the terminal, not to the agent. So, it's not going to Anthropic. All of this stays private on my VPS. Boom, paste that in. And we should be good. Check now. It's already going to run the OB sync by itself. Yet, it it ran the OB sync. I really don't have to do anything. Vault synced. And now we need to check if Hermes agent can access it. Well, let's see what's happening. Blah blah blah. Now installing the system this service for continuous auto sync. It's doing all of this by itself, guys. It's crazy. Like, I don't know what to explain. If you don't have multiple VPSs for multiple AI agents, what are you doing? Like, you can use it for anything. I'm using it to track my calories. I'm using it to do autonomous research. I'm using it to do code development. Soft like software development. So many different things. Like, preparing for interviews with podcast guests. The use cases are endless. Like, if you're don't see the use cases, it really is a poverty of imagination. You probably just lack creativity. It's not that the AI isn't useful enough. You just don't have the right setup, or you haven't figured out what the use cases are for you. All right, it says it's done. Blah blah blah, Hermes pointer to vault, check your Mac now, VPS test should appear in the Obsidian. Let's go here. And we have VPS test here. Amazing. So, this file was created on the VPS, and now we see it on the MacBook. So, I'm going to screenshot this and give this feedback to Pi. I see it. What now? Can you start Hermes on the VPS and have him check if he sees the Obsidian vault. Okay, so it cleared the terminal. It started Hermes agent, there it is. And look at it, it always does the sleep, right? So, this is one agent managing the other. Okay, Hermes is up, asking to verify the vault. It gives it the location, and it sees it. So, now Hermes agent can fully access all our files, and they're going to be instantly synced between our MacBook Obsidian and the VPS. Create a note saying Hermes hello, and let's see how fast it is. Literally instant guys. That That was instant. That was less than half a second, and I could see it synced to my MacBook. Okay, so basically, what I did is I sent a screenshot of this prompt, which again included in the second link below the video, and this will sync all of the Hermes skills with the Obsidian vault. So, Pi agent is setting all of that up. There's a lot of details that I personally don't want to handle. It's reading all the steps, and it's executing the setup, and we should be able to see all our skills. Oh, it's here. Okay, so the graph is populating. You can see very nice animation here. Uh that's Obsidian is known for, and all of the Hermes agent skills are being populated into our Obsidian vault right now. And what that means is that if we go to skills, we can actually edit them. You can see how it's structured and architectured, and it's even creating the links between the different skills to see like how they're connected, what references what. So, when we go, for example, to I don't know email you can see like we have a the skill for this and you can edit this, right? This is like much better than not having any observability. This is like clean markdown rendered with the details, code blocks, headers, quotations, like copy-pastable shell commands, right? This is a skill. This is a skill.md file that Hermes comes built in, pre-packaged with. And again, maybe you want to edit just one small step of this. Maybe you want to change something about the GitHub skill. Maybe GitHub issues, right? Here's the skill. You can exactly describe your preferences, change something, update it however you want. Of course, you can tell the agent to do that, but the in like before you didn't really have good observability on what is happening. Now, with this setup inside of Hermes agent, you can see all the skills, you can see all your notes, you can see which are connected semantically, which are close to each other, any links, and uh you can manage them very easily, very nicely in a way that's visual, interactive, and uh that just feels like the future of context engineering. So, your notes, your second brain, your your personal information, your data, private data, put it on Obsidian encrypted vault, sync it between your devices, your phone, your MacBook, your VPS, and let Hermes agent manage your second brain. This is the future. It's uh very obvious and again, the setup I just showed you all the steps in this video. So, if you haven't done them, watch it again and implement the setup. It's not that difficult. Now, to show you the actual power of this, I'm going to go back to tmux. So, we have Hermes agent here. Boom, I'm going to say uh run clear on the VPS. Now, start Hermes on the VPS. And we're going to say Actually, we're going to use the slash go feature, which is one of the most powerful features inside of Hermes agent. It started off inside of a Codex, but Hermes added it now. So, I'm going to do slash go and basically I'm going to say scrape YouTube to find the 50 highest views videos about cloud code and then fetch their transcripts as raw TXT files and save them into our Obsidian vault as separate markdown files. Let's interrupt this. Boom. So, {slash} go. And it's going to be focusing on this goal, right? It already comes with pre-built skills like YouTube content and Obsidian. And you can see the goal set. And the beauty about the goal is that you have a clear end state. If you have a verifiable, quantifiable end outcome, which in this case is finding the 50 highest views YouTube videos about cloud code and saving them as markdown files. This is not debatable, right? If you have 45, you're not finished. If these are not the highest views videos, it's not finished. If these videos are about Codex and not cloud code, it's not finished. And what that means, the beauty of {slash} goal feature, is that it works until the end outcome is finished. It doesn't matter if it's 10 minutes, 30 minutes, 5 hours, it will just keep going. And again, Codex probably has the best {slash} goal feature. Hermes agent probably the second best. But Hermes agent is a general agent. It's not optimized for just coding. You can do anything with it, right? So, now it's asking for permission. I'm going to approve that. And it's going to keep working until it figures figures this out and until it saves these files in the location of our vault. And if I had to do this manually, finding the 50 highest views videos on YouTube about cloud code, that's like hours of work. Instead, I just spoke it in like 10 seconds. I just type in {slash} goal and described literally in 10 seconds using my voice what I want to accomplish. And now Hermes agent is going to do all the steps, which would take me at least 3, 4 hours of real human work. This is the leverage that AI agents can give you. So again, if you want this setup for yourself, make sure to set up hosting your VPS. It's going to be the first link below the video. It's one of the most affordable and easy ways to set up a VPS. And everyone should absolutely be running their own virtual private server in 2026. And if you want all the resources prompts skills terminal commands, setup, everything I showed you in this video, click the second link below the video, enter your email, and you're going to get all these bundles completely for free. With that being said, thank you guys for watching. I hope you found this video valuable, and I'll see you next time.