Modern Creator
Brock Mesarich | AI for Non Techies · YouTube

Claude Cowork + Obsidian Just Changed How I Work Forever

A step-by-step setup guide for giving Claude Cowork a persistent second brain using Obsidian -- free, local, and self-updating.

Posted
1 months ago
Duration
Format
Tutorial
educational
Views
25K
541 likes
Big Idea

The argument in one line.

Obsidian solves Claude Cowork session amnesia by turning every AI conversation into a write to a local file system -- so each session compounds on the last without any extra effort from you.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…
  • You use Claude Cowork regularly and are tired of re-explaining your projects, clients, or context every time you open a new session.
  • You want a memory system that updates itself -- pulling from your calendar, inbox, and tools -- without building a custom integration.
  • You already use or have tried Notion and want something that lives on your machine, works offline, and costs nothing.
  • You manage multiple clients, video projects, or ongoing initiatives and need Claude to track all of them across days and weeks.
SKIP IF…
  • You only use Claude for one-off tasks with no recurring context -- the setup overhead is not worth it.
  • You are comfortable with a paid cloud-based knowledge management tool and do not want to change workflows.
TL;DR

The full version, fast.

Claude Cowork is powerful but stateless -- it forgets everything when a session ends. Obsidian is a free, local markdown editor that stores notes as plain files on your computer. Because Cowork can read and write any folder you point it at, connecting the two gives Cowork a persistent memory. You set up a vault, give it a CLAUDE.md map file that explains your folder structure, then add scheduled tasks that pull from Gmail, Calendar, and Slack every evening to auto-populate a daily note. The Zapier MCP extends this to 8,000 additional apps. Over time the vault compounds -- Claude knows more about your business with each passing week.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →
Chapters

Where the time goes.

00:0000:43

01 · Intro -- why Cowork needs a second brain

States the core problem: Cowork forgets everything between sessions. Promises setup walkthrough by end of video.

00:4301:34

02 · What Obsidian looks like

Visual preview of the Obsidian graph view -- interconnected nodes representing notes/documents.

01:3402:19

03 · The problem: Claude forgets everything

Enumerates the pain: lost research, scattered meeting notes, constant context re-explanations.

02:1902:31

04 · Obsidian in 30 seconds

Quick definition: think Notion but files live on your computer, notes link like Wikipedia, all readable by Cowork.

02:3103:25

05 · The brain analogy

Vault = brain, notes = neurons, links = synapses. Explains how knowledge connects to itself.

03:2504:03

06 · Why local files beat Notion

Data ownership, offline use, no server dependency. Works across Cowork, Claude Code, Codex.

04:0304:32

07 · Folder structure walkthrough

Shows desktop folder 'second brain' with subfolders: business, daily notes, goals, inbox, knowledge, templates, videos.

04:3205:36

08 · Daily notes inside Obsidian

Shows a daily note populated by Cowork with active video tasks, todos, and content plans from the previous session.

05:3606:04

09 · Linked notes and graph view

Demonstrates wiki-style hyperlinks inside markdown notes and switching to the graph view to see cross-document relationships.

06:0407:15

10 · Live demo: Cowork reads Obsidian

Prompts Cowork about current video status -- Cowork correctly retrieves four in-progress video titles stored in the vault.

07:1508:52

11 · Live demo: Cowork writes to Obsidian

Prompts Cowork to mark the Obsidian video as completed. Verifies the file moved from in-progress to published folder inside Obsidian in real time.

08:5210:16

12 · Live demo: Obsidian changes reflect in Cowork

Manually renames a video title inside Obsidian, then asks Cowork for the title -- Cowork returns the updated name, confirming bidirectional sync.

10:1611:18

13 · How to download and set up Obsidian

Walks through obsidian.md download, creating a new vault from a folder on the desktop.

11:1811:56

14 · Pointing Cowork at your vault

In Cowork, open new task, select the vault folder, click always allow. One-time setup.

11:5613:11

15 · Brain-dumping your folder structure

Gives Cowork a plain-language description of desired folders. Cowork builds the subfolder tree and populates starter files.

13:1114:58

16 · Writing your CLAUDE.md

Uses a prompt to have Cowork interview the user and generate a CLAUDE.md: who you are, folder map, how to answer common questions, style rules.

14:5815:35

17 · Testing the CLAUDE.md

Asks what am I working on -- Cowork correctly references the vault folders as specified in CLAUDE.md.

15:3516:19

18 · Supercharging with scheduled tasks + connectors

Introduces the power combination: scheduled tasks pull from external apps and write context into Obsidian automatically.

16:1917:32

19 · Live scheduled task demo

Shows a real scheduled task running daily at 6PM: pulls Google Calendar meetings, flags urgent Gmail, populates a daily Obsidian note.

17:3218:17

20 · Creating your own scheduled task

Demos creating a new scheduled task via natural language: check email and YouTube videos folder daily at 6PM.

18:1718:48

21 · Connectors walkthrough

Shows Cowork connectors panel -- Gmail, Slack, Google Drive one-click connections.

18:4819:45

22 · Zapier MCP hack

zapier.com/mcp lets you build an MCP server connecting 8,000+ apps to Cowork. Add via connectors panel by pasting the Zapier MCP URL.

19:4520:28

23 · Final second brain recap

Shows the fully populated graph view of the actual working vault. Notes it builds and improves passively over time.

20:2820:54

24 · Outro + CTA

Subscribe, join Skool community, 50+ Cowork skills available inside.

Atomic Insights

Lines worth screenshotting.

  • Obsidian is just a folder of markdown files. There is no sync server, no API, no subscription -- the intelligence is Cowork reading and writing the same files Obsidian displays.
  • A CLAUDE.md file in your vault root tells Cowork what each folder means and how to answer common questions -- eliminating repeated context explanations across sessions.
  • Changes made directly inside Obsidian are immediately visible to Cowork in the next conversation because they share the same local folder.
  • Scheduled tasks can write a daily briefing note to Obsidian every night, pulling from Gmail, Calendar, and Slack -- so you wake up with Claude already current on your priorities.
  • The Zapier MCP connects 8,000+ apps to Cowork without any native integrations -- one URL paste in the connectors panel is the entire setup.
  • Any AI agent platform that can read a folder -- Cowork, Claude Code, Codex -- can share the same Obsidian vault, making the memory system tool-agnostic.
  • A vault accumulates context over time. Each conversation that writes to it makes the next one more accurate, turning a simple folder into a compounding business asset.
  • Pointing Cowork at a new empty folder and brain-dumping your project structure is all it takes to bootstrap the second brain -- no templates, no configuration files needed.
  • Unlike Notion, Obsidian stores data locally: you own it, it works offline, and you can move it between AI tools freely because it is plain text.
  • The CLAUDE.md interview method -- have Cowork ask you questions then generate the file -- produces a map file in minutes that would take hours to write from scratch.
Takeaway

Your AI forgets. A local folder does not.

WHAT TO LEARN

Connecting Claude Cowork to an Obsidian vault turns disposable AI sessions into a compounding knowledge base that improves with every conversation.

  • Claude Cowork is stateless by design -- it reads and writes files but holds no memory between sessions. Obsidian exploits this by making a local folder the persistent store.
  • The CLAUDE.md file at the vault root acts as a standing briefing document: it tells the AI what each folder means and how to answer recurring questions, eliminating repeated re-explanations.
  • Any change made directly inside Obsidian -- renaming a file, updating a status -- is instantly reflected in the next Cowork conversation because both apps are reading the same local folder.
  • Scheduled tasks remove the manual work: a once-configured daily task pulls from Gmail, Calendar, and Slack to write a current briefing note into Obsidian every evening without user input.
  • The Zapier MCP extends connectivity to 8,000+ apps via a single URL paste in the Cowork connectors panel -- no code, no custom integrations, no per-app setup.
  • Because Obsidian stores plain markdown files, the vault is tool-agnostic: the same folder works with Claude Code, Codex, or any other AI agent that can read a directory.
  • The brain-dump setup method -- just describe your desired folder structure in plain language and let Cowork build it -- means there is no configuration barrier to starting.
Glossary

Terms worth knowing.

Claude Cowork
The Claude Desktop app run in a collaborative task mode, allowing Claude to read and write files on your local machine within a chosen folder.
Vault
The top-level folder in Obsidian that contains all your notes. Everything inside is plain markdown files; Obsidian just visualizes and links them.
CLAUDE.md
A markdown file placed at the root of an Obsidian vault that instructs Claude how the folder is organized and how to answer specific questions from its contents.
Scheduled task
An automated Cowork routine that runs on a cron-like schedule -- for example, every evening at 6PM -- to pull data from connected apps and write it into Obsidian.
Connector
A native integration built into Claude Cowork that grants it access to a third-party service such as Gmail, Google Calendar, or Slack.
Zapier MCP
A Model Context Protocol server provided by Zapier that exposes 8,000+ app actions to any MCP-compatible AI client, bridging tools not natively supported by Cowork connectors.
Graph view
A visual map inside Obsidian showing all notes as nodes and the wiki-style links between them as edges -- useful for seeing how knowledge areas connect.
Daily note
A dated markdown file automatically created in Obsidian (often by a scheduled task) that serves as a daily briefing: calendar, inbox priorities, project statuses.
Resources

Things they pointed at.

Quotables

Lines you could clip.

00:00
Claude Cowork and Obsidian is the combo I didn't know I needed.
Strong opening hook, no setup requiredTikTok hook↗ Tweet quote
15:35
This is where it gets stupid powerful.
Punchy transition line, builds anticipationIG reel cold open↗ Tweet quote
03:25
Since these are actual files inside of folders on your computer and they are stored locally, you own all this information.
Ownership argument -- speaks directly to anti-SaaS audiencenewsletter pull-quote↗ Tweet quote
02:19
Think Notion, but files living on your computer. We have specific notes that link to each other like Wikipedia.
Clean 2-analogy explanation, zero jargonTikTok hook↗ Tweet quote
19:45
This memory bank is just gonna build over time and just get better and better and know more about you, know more about your business.
Compounding value pitch -- strong closerIG reel cold open↗ Tweet quote
The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

metaphoranalogy
00:00PodCowork and Obsidian is the combo I didn't know I needed. It basically gives Cowork a second brain, and it has hands down made one of the biggest impacts on my entire system. Because if you've used Cowork for a while, you already know this problem.
00:13You do amazing work inside of it, whether that's research, client context, or planning, then you open it up tomorrow and have no clue where any of it went. And Cowork doesn't remember a single thing you talked about yesterday. Well, Obsidian fixes that.
00:25So in this video, I'm gonna show you what Obsidian is, why it instantly levels up the way you use Cowork, and exactly how to set it up the same way I did. And the best part about this is it's completely free. It lives on your computer, and once it's dialed in, it can run on autopilot while you sleep.
00:41So without further ado, let's get into it. Alright. So before I break down what exactly Obsidian is and how you could begin leveraging it, let me show you exactly what it looks like, and it looks pretty crazy.
00:51It's definitely some eye candy here. It's pretty cool to look at. So this is basically our second brain that we are forming.
00:58This is basically taking all this context and information from Claude Cowork and basically giving us a visual here in a place to store all this information as context, and it functions very similar to a brain, which we're gonna talk about later in this video. But if we look at this, this right here says school overview, and this is connected to a bunch of different documents here.
01:16And we could basically see all the connections between these different documents. So if I click on this, it will then open up this file here, which is basically a breakdown on my entire school community. So think of this as, like, a neuron that is trained on my school community.
01:30We're gonna dive deeper into the interface, but let's quickly now talk about exactly how this works. So I honestly think that everybody nowadays wishes they could a second brain, a place where they could dump all this context and information. Think about things like OpenCLOB.
01:43People are trying to use something like that as their second brain. Not only is it a place to work, but it's a place to store all this context and information that we give these large language models like Claude and specifically Claude Cowork. So this right here probably sounds familiar to you, and Obsidian is gonna solve this problem by the end of this video.
02:00So for example, you did great research last week, but you have no idea where that went inside of Claude. Meeting notes are scattered across dozens of old sessions that you have. You keep having to reexplain the same context every few conversations, and then you could even ask the question, what do I know about this client?
02:16And there's no way to get an answer to that because Claude just simply forgets. So to explain Obsidian in thirty seconds, here's exactly how I would break it down. So think Notion, but files living on your computer.
02:28We have specific notes that link to each other like Wikipedia. So everything that you see right here that is in this brain like interface is just a markdown file inside of a specific folder. And this means that Claude Cowork can read and write to all of it, and that is how we store this information here.
02:44It is free. It's local. You don't need to set up any APIs.
02:47I'm gonna show you exactly how to do that later in this video. It's really simple to do. So you might be asking yourself, how is this a second brain?
02:54So let me break down quick analogy for you right now so you could better understand it. So our vault, which is basically our folders in our work space inside of Obsidian, is basically just our brain.
03:05Each note is a specific neuron as you can see here, whether that's a meeting note, research, specific idea, goal, daily note, client note. All of these different things are neurons for our brain. And then think of the links between them as a synapses.
03:18And with this, our knowledge connects to itself, and we could basically see the different connections we have to each of these different documents and memories. So let's just break down the basics here so you could get an understanding of what Obsidian actually is at the core.
03:31A vault is basically just folders that are stored on your computer just like when you are using Claude Cowork. And the amazing thing about this is this is not stored on their servers. So if you're using something like Notion, that is stored on the Notion servers.
03:46You have to be on the Internet to actually pull up these files. Since these are actual files inside of folders on your computer and they are stored locally, you own all this information. On top of that, it's 100% free and works completely offline.
03:59And, again, CoWork just opens up this folder and can read and write everything. Not to mention, if you want to use this between CoWork, Codex, ClaudeCode, all of these different AI agent platforms, you can do so because it's all stored within a folder.
04:13That's basically as simple as it is. So to show you exactly what I mean, this is a folder I have on my desktop. It is called second brain.
04:20And inside of that, we have multiple different subfolders. We have one for business. We have one for my daily notes.
04:26We have one for my goals, inbox, knowledge, templates, and videos, and plenty more inside of that specific folder. So at the foundation, this is what this looks like.
04:36But inside of Obsidian, it is structured like this. And to show you exactly what I mean, let me go inside of my daily notes folder inside of Obsidian.
04:45So if I come over here on the left hand side, we are going to see basically like a menu of a couple of different things. These right here are those subfolders that I just showed you. So I could click on my daily notes folder, and I could, for example, pull up one of these.
04:58This is my daily note for today. And as you could see, it has a couple of different things. And a key thing I want to note is this looks very similar to Notion, which is a software I've been using every single day to help run my business over the last three years.
05:11But now instead of me needing to use Notion, I could just automatically use it in Obsidian, and I could make changes here manually. As you could see, I'm typing here, or I could just have Claude Cowork go and make these changes on autopilot since this is basically just markdown files. It has context on what exactly I'm working on, so it knows the videos that I need to actually film today.
05:30It knows that I need to write a school community post. All these different things here, has context on so it knows what I'm working on. Because I was going back and forth with Claude Cowork yesterday, and it saved this context inside of this file.
05:41On top of that, we're gonna see different hyperlinks inside of our notes. So if I click on this, it will take me to another note, which is basically a video that I'm working on, which is this one.
05:52So you could see the core concepts, why we're filming this video, key demos to show, um, and even shows potential titles. And from here, I could always switch back to the graph view to see how all of these different files are communicating and reference each other. Alright.
06:05So let me go ahead and show you this live in action so you could see that this is working. So what we're gonna do is we're gonna come over to the Claude desktop app. If you don't already have it, make sure to download it.
06:15There will be a link in the description to do so. Once we are inside of here, we want to make sure we have co work mode toggled on inside of Claude. And then from here, we need to actually select a folder.
06:25And this part is really, really important because this is basically how we're going to sync this up with Obsidian. So for me personally, I named this folder second brain. You could name it whatever you want.
06:36I'm gonna click on allow. And in order for me to see if this is working, let me just give it this quick prompt. Where am I at with all the videos that I'm working on?
06:44The reason I'm asking this question is because yesterday, I was doing back and forth with Claude Cowork and basically explained all the videos that we were working on. So I had them stored inside of my YouTube videos project here that we could actually see inside of Cowork right here. So there's a couple of different videos that I've been working on, and the way that it's able to reference this is because this information was saved inside of that second brain folder that we are using inside of Obsidian.
07:09Whenever there is a change in Cowork, it automatically gets added here because these are the same exact files. They're just in a different view. Alright.
07:16So just like that, it pulled up these four different videos that I am working on right now. We have 10 Claude co work skills you didn't know existed, seven Claude co work skills every marketer needs, AI second brained, which is this video, cloud managed agents. And in order for me to double check this inside of Obsidian, let me pull that up to see if we could reference these.
07:35So I come over, we come to YouTube videos, and I click on in progress. You could see we now have these four different subfolders inside of Obsidian.
07:44If I click on them, you could see the outline for each of them. Alright. So let's try something out.
07:48I want to give a prompt and make a change inside of CoWork, and then I wanna see if it's reflected inside of Obsidian so that way we could know that our second brain is functioning properly. So let me go ahead and do that now. Hey.
08:00I want you to go and change the Claude CoWork Obsidian video and mark that as completed as we finished filming this, and that's ready to be posted tomorrow. So we're gonna give it prompt. It should be making changes to this file, and it should be reflected inside of Obsidian.
08:15But let's go ahead and check it out in a second. Alright. So according to Claude, this is done, but let me come back inside of Obsidian.
08:21If I come over to YouTube videos and I click on in progress, you can now see that the Obsidian video is no longer in progress. It should actually be in the published folder.
08:30And as you can see here, that is correct. The AI second brain Obsidian file is now in the published, so that is accurately reflected.
08:39And then to show you what this looks like inside this actual folder on my desktop, we could see the exact same thing. So we have YouTube videos. We could see published, and we now have the AI second brain obsidian right here added to that.
08:52Alright. So in order to show you that changes can be reflected from both sides, from the Obsidian side but also the CoWork side, let me come into Obsidian, and let's actually just change a folder here. So I'm gonna open up videos that are in progress.
09:03Let's open up the 10 skills SQL video. So as you can see here, we have this markdown file. We have the specific status, and it shows planning.
09:12Priority is high. The topic is skill sequence. We even have these tags here as well as the date that it's created.
09:18So what I'm gonna do is let me just come here and change the title for this. So let's call this 10 clogged co work skills I can't live without, steal them. So I'm just gonna call this title.
09:29So now we know that this is gonna be the title for the video. And since we made this change here, I'm gonna come back to co work, and I'm gonna simply say, tell me about the video that we're working on about the 10 Claude skills. What is the title for it?
09:41Since we just made that change in Obsidian, again, since this is just a markdown file and we're working inside of the same folder inside of both of these platforms, we will then see this accurately reflected here. We should have the new title here added. And here we go.
09:56We got this response from Claude, basically saying that the video is in this folder. Here's a rundown. Here is the title.
10:02So that information is accurately getting reflected from Obsidian to Claude so we know that any of the changes we make directly inside of here will be reflected into Claude because, again, these are just files that are stored inside of the same folder. Alright. So I'm gonna move on and show you exactly how you could download this and set this up yourself inside of Claude Cowork.
10:23So first of all, we're gonna come to obsidian.md. There will be a link in the description to download this. It's completely free to download.
10:30So what we're gonna do is click get Obsidian for Mac OS. It's gonna begin downloading in our browser here. Let me go ahead and open this up.
10:36And what we're gonna see here now is a couple of different things. So we could create a new vault. We could open folder as a vault or open vault with Obsidian Sync.
10:45So there's couple things you could do right now. You could go ahead and add a new folder to your desktop, or you could open an existing folder you already have. But for the sake of this video, I'm just gonna create one from scratch.
10:56So I'm just gonna open up a new folder on my computer. I'm going to name this folder Obsidian, and then I'm going to open folder as a vault. And all I'm gonna do is look up that Obsidian folder here in Finder.
11:07And as you can see, I can click on this and open it, and it's an empty folder. There's absolutely nothing in here. If I pull it up, you could see that there is nothing in there.
11:15And now we have this blank interface inside of Obsidian. Next up, we need to come back to Cowork and actually point this at that folder that we just created on our desktop. So what I'm gonna do is I'm gonna come up to co work, open up a new task, and then I'm gonna click on this folder here.
11:32And I'm gonna scroll down and click on choose a different folder, and then just look up Obsidian or whatever we just named our folder there. For example, let me just click this and open this. I'm gonna click on always allow for it to make changes.
11:43And just like that, we are now working inside of this. So what I'm personally gonna do is just give a brain dump on everything I want this to have context on so we can actually form this second brain with different subfolders and categories. Alright.
11:57So I want to actually form this as my second brain inside of Obsidian. What I want you to do is add a couple of different sections. I want you to add a folder called YouTube videos, and this is where I basically have everything YouTube video related.
12:11This is, like, for all of the prep for my videos. I want you to keep track of videos that are in progress, videos that have been published, and then even keep track of the performance of all these videos. I also want you to create a separate section in this folder, and that's gonna be for school community.
12:26So this is everything in regards to my school community, as well as I want you to add a section here for all of my client work. So I'm just gonna send that off, And what this is gonna do is it's gonna create those different subfolders inside of that folder, and it should be populated inside of Obsidian. So let's just give it a second.
12:43Alright. And just like that, it now says that our second brain structure is set up in Obsidian. Here's what it built.
12:47It added the YouTube video central hub. Added school community, and added client work. So if I come back over to Obsidian, you can now see this is accurately reflected here.
12:57So we have client work. We have active clients, archive. We have prospects, templates, all of those different subcategories.
13:03And inside of school community, we have all these other things. We have content. We have courses.
13:07We have events. All that stuff is now structured inside of Obsidian. Next up, we need to talk about the Claude MD file, and this is basically going to be the map to the brain.
13:17So you probably already know what a Claude MD file is. This is basically something that's loaded every single conversation, instructing Claude how it should operate.
13:25So that way, you don't need to keep repeating yourself. So this is basically explaining who you are, how you want it to perform, and very general instructions on how Claude should be interacting with you. So for a second brain, this is super important because this tells Claude Cowork what each folder means and how to answer specific questions from your vault because we're gonna have lots and lots of context inside of our second brain, and it needs to understand how we could best take advantage that.
13:50Alright. So next step, what we need to do is actually create this clotMD file. So what I'm gonna do is I'm gonna use this prompt.
13:56I want you to write a clotMD file for the root of this vault. This is my second brain. Basically, it says, look at my folders I've set up.
14:02Ask me any quick question you need about who I am, what each folder is for. Keep it tight, about 30 lines. So, basically, it's gonna have four sections, who I am, folder map, how to answer common questions, and style rules.
14:14So what this is gonna do now is it's gonna interview me in order to populate the clotMD right here, which is basically the instructions for this vault. Alright. So it's gonna take me through a series of different questions.
14:25So I'm not gonna bore you. I'll go through all of these and then show you the final outcome of the clotMD file. And just like that, it generated the markdown file.
14:32What we could do is we could click on this here. Basically, it breaks down who I am, shows the folder map, so YouTube videos, school community, all that stuff. And then it even shows how to answer common questions.
14:42So if I ask what am I working on, it basically instructs it to check the YouTube videos in progress folder plus the client active deals dashboard. So it's basically just instructing Claude where to look inside of the Obsidian folder in order to get specific information. And now to test this out, let me just give it the prompt.
15:00What am I working on? And what it's gonna do is it's gonna reference those different folders that we just set up and instructed inside of the CloudMD.
15:08So it said that we're working on these five. That is great. And this is just, like, populated with fake information.
15:13This isn't actual client deliverables. Since I just created this new folder for you guys, I needed to add specific context in there. Now if we go back to Obsidian, that we have a more populated, you know, folder.
15:25So we have client work. We have different active clients. We could take a look at the overview here.
15:29All this information now is accurately mapped from, you know, what we have created inside of cohort. What I'm about to show you is where this gets really powerful. And what I'm talking about is when we combine scheduled tasks with connectors, MCP servers, and Obsidian.
15:44So this way, we're not only creating this brain, so to say, by going back and forth with co work, but it's actually pulling all of this different context from all of your different applications on autopilot and automatically compiling this into Obsidian, otherwise known as our second brain.
16:00So if we add connectors to Cowork, our brain is able to see basically everything that we can see. So it can connect to our calendar, Gmail, Slack, Google Drive, basically any of the different applications that we use. It's able to pull context from and store that in our brain.
16:13Instead of just talking about scheduled tasks and how powerful they are, let me show you it live in actions. So I created this one that runs every single day at 6PM. Not only does it pull from my Google Calendar and check what meetings I have, it also pulls from my Gmail to basically flag all the urgent emails so that way I could respond to them in the morning, but it also populates the daily note inside of Obsidian that we actually create.
16:37So if I open up this markdown file, you could see what's on my calendar for tomorrow. It breaks down the top three things that I need to do tomorrow. It breaks down the priority inside of my email inbox, and then it even shows the active videos I have in the pipeline.
16:51And this right here is all pulled from my Obsidian second brain. So let me show you exactly what I mean. If I come over on the left hand side and look at my daily notes, you could see we have this new one that was just generated with this scheduled task.
17:04It automatically creates this for me every single day. And inside of here, we could see everything that I just showed you. It shows a calendar.
17:10Shows our top three things for the day. It shows everything in our email inbox. It even shows our action items that we need to move forward with.
17:17And then here are those videos that we are working on that were inside of the video pipeline. It even shows the hyperlinks so I could pull up this separate file, which is another markdown file inside of Obsidian, as well as we could click on this and see each of our active clients.
17:32So let me show you exactly how you can do this exact same thing. So what you do is basically come to new task, open up that Obsidian folder or whatever folder you're working inside of, and then basically just explain what it is you wanted to do. For example, let me show you what I mean.
17:45I want you to go through my email inbox every single day as well as look inside of the folder about my YouTube videos and let me know what videos are in progress, what their status is, and then break down all the emails I need responding to tomorrow. Run this as a scheduled task every day at 6PM. And just like that, we have a new scheduled task that runs every day at 06:09PM.
18:04We could actually click on this, see the instructions here, or we could always see all of our scheduled tasks on the left hand sidebar and see all the ones that are running. Now that's all great, but we need to connect these to the different apps that we're using. So let me show you exactly how to do that now.
18:17We're gonna come back to co work, and then what we're gonna do is we're gonna come down to customize, come to connectors, and we're just gonna pick all the different apps that we use. So for example, if I use Gmail and wanna pull from my emails, I would click on this and basically just connect it.
18:30Same thing for Slack. If I wanna connect Slack, I would just come here and click on connect. And then let's say there's some specific app inside of Claude that you can't connect to.
18:38Well, I'm gonna show you a hack that will instantly connect you to 8,000 plus different applications so you can basically fill your second brain with context across all these different apps that you're using every single day, and this is possible thanks to the Zapier MCP. So we're gonna come to zapier.com/mcp. There will be a link in the description to use this.
18:57We're gonna click on start building. And from here, we're just gonna create a new MCP server. So I'm gonna click on that, then I'm going to select our client as Claude Cowork.
19:07And from here, you could see I have some different applications already added here. I have school. I have HubSpot, Stripe, a couple of other ones.
19:14And in order to add a app, we just come here and just search for any of the apps we use. So let's say we want to connect to HubSpot. We could go ahead and click this, and then there will be a button here that just says connect.
19:24And then once we connect that, we will have access to all these different actions across that entire app. Then in order to connect this to Claude Cowork, we're just gonna click on connect, click this add to Claude button. And then from here, we're gonna come back to our connectors.
19:37We're gonna click on browse connectors, then look up Zapier. And all we're gonna do is paste in that URL that Zapier MCP gave us.
19:46And just like that, guys, we now have a second brain that we could begin populating and building upon. Right now, mine looks small because I was demoing this for the video.
19:55But if I go back and show you one that I'm currently using right now, you could see that this is basically mapping out all of the different documents that I create inside of Claude Cowork and whether or not you use the interface inside of Obsidian because you probably might not use it much. You're not gonna come in here and individually edit these files.
20:12You could just have Claude Cowork do it, but you will really notice the power of this as you're pulling context from previous conversations or specific files inside of Claude Cowork over time. So this memory bank is just gonna build over time and just get better and better and know more about you, know more about your business.
20:28Alright, guys. There we have it. That is the Obsidian second brain that you could plug into Claude Cowork.
20:33If you guys got some value from this video, leave a like. Subscribe to this channel for more content. And if you guys wanna dive deeper, make sure to join my school community.
20:40There will be a link in the description. Inside of my community, I give away the 50 plus Claude co work skills I use every single day, and I'm building out this directory more and more every single week. With that being said, guys, thank you so much for watching to the end, and I will see you guys in the next video.
The Hook

The bait, then the rug-pull.

The loudest problem with Claude Cowork is not what it does in a session -- it's what disappears when the session ends. Brock Mesarich's answer is a folder of markdown files that costs nothing, lives on your machine, and lets Cowork read and write its own memory on autopilot.

Frameworks

Named ideas worth stealing.

02:31model

Vault-Neuron-Synapse Analogy

  1. Vault = brain (the folder)
  2. Notes = neurons (individual markdown files)
  3. Links = synapses (wiki-style [[]] connections)

A mental model for explaining Obsidian to non-technical users by mapping it to brain anatomy.

Steal forAny explainer for local-first or graph-based knowledge tools
13:11list

CLAUDE.md Four-Section Template

  1. Who I am
  2. Folder map
  3. How to answer common questions
  4. Style rules

A ~30-line CLAUDE.md template generated by interviewing the user, then loaded every Cowork session as permanent context.

Steal forAny Claude project that needs persistent persona/context without re-prompting
15:35model

Brain Sees Everything Stack

  1. Scheduled tasks -- run on autopilot
  2. Connectors -- pull from Gmail, Calendar, Slack, Drive
  3. Zapier MCP -- reaches 8,000+ additional apps
  4. Obsidian -- catches all of it in the daily note

The four-layer automation stack that turns the Obsidian vault into a self-updating intelligence system.

Steal forAny set-it-and-forget-it AI workflow architecture pitch
CTA Breakdown

How they asked for the click.

VERBAL ASK
20:28product
If you guys wanna dive deeper, make sure to join my school community. Inside of my community, I give away the 50 plus Claude co work skills I use every single day.

Soft sell after content is complete. Leads with value (50+ skills) before the ask.

MENTIONED ON CAMERA
Storyboard

Visual structure at a glance.

open -- product logos on screen
hookopen -- product logos on screen00:00
Obsidian graph view preview
promiseObsidian graph view preview00:43
Obsidian in 30 seconds slide
valueObsidian in 30 seconds slide02:19
Cowork reads vault -- live demo
valueCowork reads vault -- live demo06:04
Download and setup walkthrough
valueDownload and setup walkthrough10:16
Stupid powerful -- automation slide
valueStupid powerful -- automation slide15:35
Populated graph view recap
valuePopulated graph view recap19:45
Laptop CTA shot
ctaLaptop CTA shot20:28
Frame Gallery

Visual moments.

Watch next

More from this channel + related breakdowns.

Chat about this