Claude + Obsidian Just Changed AI Agents Forever (FREE!) 🤯A

Claude with Obsidian is the most powerful AI memory system I've ever used. And in this video, I'm going to show you exactly how to set it up. Imagine if your Claude remembered everything, every goal, every decision, every project, every person you work with, and not just Claude, but every single AI agent you use, all powered by the same brain. I built this system for free and it's completely changed the way I work.

And in this video, I'm going to show you how to give Claude a second brain using Obsidian. So every conversation you have starts with full context, smarter answers, and zero time wasted re-explaining yourself. The setup is free, the results are unreal. And by the end of this video, your Claude is going to know everything about you, your business, your goals, and your life.

And every single AI agent you use is going to become a hundred times more powerful because of it. And there's one thing I'm going to show you later in this video that blew my mind the first time I saw it. So stick with me to the end because that's what makes this whole system grow smarter every single day, completely automated. Let's get into it.

Today I'm going to show you a really powerful system that I've built into Claude, and this is Obsidian, right? And so Obsidian and Claude work together as a really powerful memory system that maps out everything about me, my life. It takes notes on me all day long, and then it plugs it into a knowledge system that all links together. As you can see right here, it's beautifully color coded.

These go directly into Claude and all my other AI agents, right? And so this is a system that uses Claude with Obsidian, right? So if we go to the memory section, you can see I've got Claude over here. I've got free Claude as well.

And both of them are plugged into my memory system. And you can see that we have this beautiful graph here with everything mapped out together, everything linking together here. And basically this understands me, my life, my goals, every person that I work with, every tool, every business I have, and it all links together inside a system that AI agents can easily map out and use together, right? And the other cool thing about this is it has memory, so you get updated every single day.

So if we go to our memories vault here, you can see all the new memories that we have. And if we go inside Obsidian directly, this knowledge vault allows you to track everything in one place. So you've got this beautiful knowledge graph here. And then if we go to our base, we have a full structure that's really neatly organized.

I didn't write any of this. My agents do it for me automatically. And then you can see, for example, it takes notes on me. It organized everything in one place, even for example, organizes everything that I'm working on.

And it just has an awesome system noted down here, as you can see, right? And so it's all tagged. It's all color coordinated and it's just the best possible system that I think you can give your agents based on what I've tracked. So let's get straight into it.

If you have any questions as you go along, if you're watching this live, feel free to ask as I go along. But basically how this works is Claude has a memory and Obsidian gives it a brain. That's how it works, right? So Claude over here with the memory.

And then the problem is with Claude, like it remembers fragments, right? It might remember, for example, a name, a preference, what you told it last week. It has no idea what you actually worked on, what decisions you made or what's sitting on your plate. So this actually fixes it.

This system fixes it. So you can wire Claude to Obsidian as a second brain, capture context automatically with something called OMI, organize it with something that's called the para structure. And every agent you run after that becomes a hundred times more powerful because it understands you, everything about you and everything else, right, and it maps it into this beautiful knowledge system. Another cool thing about this is like all the tools that we're using are free, right?

OMI is free and Obsidian is free. And so there's three pieces in the stack, but this makes your AI agents a hundred times more powerful. And the thing to know about this is like Claude's memory is pretty weak, right? That's the whole problem.

And anytime you're not using Claude, it doesn't know what you're doing. It's native memory only holds onto fragments. It doesn't see what you're working on the rest of the day. It has a few preferences.

Sometimes the last thing you said, but the moment you start a new session, it doesn't really understand what you worked on properly yesterday. What decisions you made last week, what's currently on your plate, et cetera. So you send the first, the problem is like you spend the first five minutes of every conversation, re-explaining yourself and then Claude will give you generic answers because it doesn't know your situation. And then if you multiply that across every agent that you use, Hermes, OpenClaude, Codex, Antigravity, FreeClaude, Code, you might be losing an hour a day just to re-explain in context, right?

And so these, with these AI models, I actually don't think the model or the API is a bottleneck. I think the context is. So this is a three step layer. So what are the layers here?

We actually use something, it's an open source project. It's called Homey, sorry, Homey to generate content. And this basically sees what I work on all day, right? If I ever get distracted, it actually tells me, it records my memory.

So it logs these automatically. And then also we plug that into our apps over here, right? So if we go back to memories here, you can see that it actually talks about my agent operating system here, which we were just working on a minute ago, right? So it tracks things like every few minutes and understands exactly what I'm working on.

And then we can export that to any of our apps. But the main one that I sync it to is Obsidian. You can see it's taken 1,800 memories, it stores it locally here. We can sync at any time and we can use Homey for doing that, right?

So Homey is a capture layer, right? It's an app that listens to your day, takes notes on what you said, records your conversations, identifies tasks, and it's free to use as long as you skip the only chat subscription, which you don't need, and then it just captures your context all day, right? Pretty amazing stuff. Even for example, creates tasks for you as well, as you can see right here, you can rewind, it can actually, if you want it to, it can record your conversations and what you're saying on the microphone, which is pretty crazy.

And then also it can screen record too, right? So it can screen capture and that sort of thing, which is pretty amazing. And so what that does, that step one, it plugs it over to Obsidian, right? As you can see right here.

And that's why we have such detailed notes. And the other thing about this is you're not doing this manually. You don't want to be doing this manually, right? That's a lot of work, a lot of work.

So what do you do instead? Instead, what you want to do is you want to have your AI agents trained to take notes on you, right? Which is what we do with Homey. So let's see what questions we've got here.

And by the way, feel free to ask any questions as we go along. So Trad said, Claude re-upped his tokens today. Good stuff, man. And then Fabricant says, does the Homey's SEO system, is that what I'm using with my websites?

Yes. Yes. Fabricant says, Homey is subscription or open source? It's open source, right?

So again, if you don't use the chat inside Homey, it's free to use and it will still take notes on your day. So that's how you can use it. You don't need to use a subscription. You can use it for free.

All right. So you're Obsidian and this is the vault. So it's basically everything inside Obsidian is written in Markdown. So if you go to our base over here and you see all of these notes that we've been taking, for example, like this.

And basically what this is doing is it's taking notes on like everything that I work on, all the people, everything else, right? And it creates this tag system with all this information about me, what I'm doing, what I'm using, everything else, right? And so this is basically a way to organize in Markdown files, all your folders and text files on your machine, right? And so it's simple.

It's just like files and folders. It's nothing crazy, but it turns it into a beautiful knowledge system. And I always think like the simplest systems are the best. So Obsidian is a vault, stores all your knowledge, stores all the notes.

Claude can read from it. Other agents can read from it. Your second brain lives here forever. And you own it because it's set up locally, right?

And then finally you have Claude, which is the reader and the organizer. Now this is pretty interesting because what I do inside Claude is I actually have a pinned conversation for Obsidian. And anytime I want to improve it or organize it, or I need to fix something, I can go into Claude and have this conversation called Obsidian, right? And so I can just speak to Claude.

It organized my notes. If I need to, you can make it better. I can organize it better. And it's basically the reader and the organizer, right?

And it has two roles. So number one is it reads a vault for full context before answering. And number two, it organizes a vault into a useful structure. So both happen automatically.

If you combine OMI, Obsidian, and Claude, three together, you get a second brain that captures itself, organizes itself, and feeds every AI agent that you run. Right? OMI captures, Obsidian stores, Claude reads and organizes. So that's basically it.

And that's how it works. OMI is a listener, captures everything all day. Obsidian is a free note taking app, right? It just stores plain markdowns files in your folder.

It's just a folder on your machine. But it basically creates a map of all these different notes of your life, which is amazing. And the interconnection happens when you link one note to another note, right? That's how Obsidian is building the knowledge graph.

So if you go back over to this knowledge graph here, you see how these are all linked? It's because we basically have these tags that link everything together. So if something is a tool, my AI agents will mark it as a tool, right? As you can see right here, right?

It even takes notes on me, who I am, my personal life, tags it, as you can see here, and then that helps create a beautiful knowledge graph, right? And then we have these settings for color coordinating it with groups. So if it's underneath people, if it's categorized as a person, it's going to be this color. If it's, for example, a tool, it's going to be marked as this color, et cetera.

And why does this matter? Every AI agent in 2026 speaks markdown, right? Claude reads it natively. So does Codex, Hermes, Antigravity, OpenClaw, PreClaude, Code, Gemini, right?

Your vault becomes a universal memory layer for every model that you use. And so the interconnection happens because you link another note to another note, Obsidian builds a graph view that maps the relationships, and that's how the second brain becomes a brain. A brain is just a network, and that's what we have right here, rather than just a stack of files. And then finally we have Claude.

Let's see what other questions we've got here. Fabricant says, one more question. Honestly, feel free to ask any questions. One more question.

How do we make the five website strategy? Yeah, so you would have five different websites and then you post SEO content to them. That's how it works. And then you can link them to Claude or your AI agents, basically.

All right. And then we have Claude, which is the reader, right? And this is two ways to wire Claude to your vault. Then you can pick whichever one fits for you, right?

So you've got MCP Obsidian, which is cleanest. So you can, for example, install Obsidian MCP. This is an open source project that links Claude and Obsidian together, right? And that'll work in the cloud because it's an MCP.

MCP is just a way to link two things together, like Claude and Obsidian. And then you've got a folder, right? So you can just have a local folder. This is what I actually do.

It's simple. You don't need to install anything. It's just good for trying out. And once wired in, you can ask.

Check for my memories and surface some ideas for what I could automate based on what I do day to day. Right. We could plug that into Claude and Claude is linked to Obsidian and we're good to go. Right.

And that single shift from generic answers to answers grounded in your real context, that is how you make this a hundred times better, right? Because every improvement layers on top of it. So you can have Hermes, you can have any APIs like our alpha, you can have open core and they're all linked into this system. And then the final step here is using Claude to help you organize it.

Right. And this is what most people miss. So after Claude reads your vault, you can ask it to organize it. So a prompt you could say is based on Obsidian vault and you restructure it into a useful second brain using para, create maps of content so the navigation is useful, not just tidy and Claude can restructure the whole vault.

So for example, mine went from like loads of untitled documents and one giant memories file to a clean structure with daily notes, maps of content, SOPs and a graph view that links it together. Right. As you can see right here, it's a beautiful graph, right? Beautiful information right there.

And you might be wondering, okay, what is para? So this is Tiago Forte's structure for second brains. And you have four buckets, right? Claude already actually understands it.

So you don't need to teach it the rules. Basically you have projects, right? So active work, for example, won't want to work on this quarter. Once it's completed, he'll move to archive.

Then you have areas. So ongoing responsibilities, for example, like that could be marketing or clients you work with, and that lives in areas. Then you have resources. So these are topics of long-term interest, right?

That could be reference material. It could be articles to save. It could be books you read, frameworks worth keeping, for example. And then finally you have the archive, right?

Everything that is no longer active inside here. So for example, that could be old projects. It could be outdated resources. It's still searchable, but it's going to be out of the way.

And so you see how this works. You've got projects, your areas, you have resources and you have the archive. So four different ways of connecting everything together, which is amazing. And so this is a flow that works together.

And the more you give your AI agents, the more useful they become. I actually call this the infinite context engine. And you can even have, for example, something like notebook LM on top of it to make it even more powerful. But yeah, this is a really cool way to organize it.

And the system isn't read-only as well. So it's a two-way loop. It's a positive feedback loop because you have the vault that goes to Claude, your AI agents, and every chat starts with full context. Claude pulls in relevant notes, surfaces what's on your plate, references your actual decisions.

But then as a feedback loop in another direction, you've got Claude that updates the vault, right? So every chat writes back useful insights, notes on the conversation, action items, decisions, new resources. So the vault grows and improves. So the more you use it, the smarter it gets.

Every conversation feeds to the next one. And by month six, your second brain has more useful content on your business and on your life than you yourself can remember, right? That's the crazy stuff. So the more you use AI agents, the more useful they become.

And bear in mind, the setup is free. The only thing blocking you or stopping you from doing this is probably a belief that's holding you back, right? Now, we have loads of people who are willing with AI inside the AI profitable community. So I know that anyone can do this.

I know that if you don't have any experience, that's no problem, right? You can see all of these different testimonials and wins from people getting awesome results right here. And so some people are going to say, I'll just use my Claude's native memory and that will improve over time. And I would say, maybe you could do that, but it's siloed off in one product, right?

And the moment you switch to Gemini or Codex or Hermes, the memory is gone, right? Obsidian as a vault actually belongs to you and every agent you ever use can read it. That's way better. Other people say setting up a second brain, that's too much work.

But the truth is setting up takes one weekend. You could do it today. Maintaining it doesn't require any effort because Omni captures it whilst you're working. Claude organizes it whilst you sleep.

The first week is the only week you actually work on it. And every week after it works on you. And then the final thing is some people say, I don't have enough notes for this to matter. I actually think you probably have more than you think, right?

For example, like your Slack history. It could be emails. It could be voice notes. It could be journal entries.

It could be old documents. Omni captures everything from one week, right? And so you could hand them to Claude with the power prompt that I just talked about and watch what comes out. And the vault just grows fast once you start.

We actually have a 30 day roadmap inside the AI Profit Boarding for implementing this. But just to recap here, Claude alone is impressive. Claude wire to a vault that captures your life and organizes it is a different category of tool, right? The setup is free.

The maintenance is automatic and every agent you ever use will pull from the same memory. That's what you want, right? That's what we have right here. That's how we can get the most out of our AI agents.

Omni captures, obsidian stores, Claude reads and organizes, Para gives it the structure and the two way loop makes it improve every day. So it's one weekend of setup. And then after that, it's amazing, right? Now, if you want to get the full Infinite Context Engine setup, the MCP install, the Para template, the hundred prompts I use daily for this sort of stuff, the SAP for organizing the vault, a 30 day roadmap, plus how to wire it into a beautiful agent operating system like this, you can get it all inside the AI Profit Boarding.

Link in the comments description or go to theaiprofitboarding.com. And you can see that we have basically this amazing community where you can get all the notes on how to set this up. We have something called the Infinite Context Engine. You can see it over here with a video tutorial and the step-by-step instructions inside the community.

You can ask any questions. I personally answer them with videos. And then also we've got a full community of 3,100 people working on similar things, and that means that you can build this out yourself, right? Which is pretty cool.

For example, like Dr. John Meza, he actually built out the full agent operating system himself using the training that we gave you. So there's just so much cool stuff going on right here, right? Inside the classroom, you get all of my best trainings.

If you're a complete beginner, you can go from beginner to expert over here. We have new daily updates, like you can see, including the agent operating system with a full zip file and all the prompts for using this stuff. We have four weekly coaching calls where you can ask questions, get help on your setup and meet other cool people doing similar things. And then also you can see inside the map, you can connect with people in your local area who are using stuff like this as well, right?

Which is pretty amazing. So feel free to get that link in the comments description or just go to theaiprofitboarding.com. So we have a question here, which is someone says, interesting. So I have Obsidian, but you are suggesting I need Claude to link it and keep context.

Does this solve my Hermes prompts and context issues? Yeah. So I think you could use Hermes to organize it as well. I like to use Claude.

I think a lot of people who watch these resources probably use Claude as well. And basically the whole point is that you can get Claude to organize it for you. Right? So if you just have the basic memory system from, if we go to our notes here and I'll show you what Omni creates for you as a memory system.

So this is what Omni creates, right? And it's just these little bullet points. But the thing is, if you then get Claude to organize that for you, which you can see over here, then it will create like a much more organized, much more in depth, whole setup and whole system for this, right? And so that it's beautifully organized.

It's more nicely formatted, et cetera. So that's why I recommend using Claude. And it also creates a positive feedback loop because you've got Claude that improves Obsidian and then Claude actually reads from Obsidian to create better outputs in the future. So they link together.

And then Fabricant says for four blogs, you have four websites. Yes. So if you want to rank four websites, like we do, we have five websites inside our SEO tool inside the agent operating system, which you can see over here. Then you would have five different domains linked, like you can see.

Fabricant says it only takes one hour to set up Obsidian with Claude. It's very easy. 100% agree with that. I'm so sorry.

I meant that I have Hermes and Obsidian, but you're saying Claude helps or no, you could change that with Hermes. It could be Claude with Obsidian, whatever you use day to day, stick with that. So if you're already using Hermes, you can get Hermes to organize it in the same way. That's it.