The argument in one line.
You can connect NotebookLM's free 300-source research platform to Hermes agent via a browser-cookie skill, then extend it with n8n to automate research queries and email workflows entirely from Telegram.
Read if. Skip if.
READ IF YOU ARE…
- A solo builder or creator who already uses the Hermes Telegram agent and wants to plug a free 300-source research layer into it without paying for RAG infrastructure.
- An n8n user who wants Hermes to build and manage automated workflows — like daily email digests — from inside Telegram instead of touching the n8n UI.
- A knowledge worker who uses NotebookLM for deep research and wants to query their notebooks hands-free from their phone while away from their desk.
- Someone comfortable with browser-cookie authentication who wants to wire unofficial Google integrations into their personal AI agent stack.
SKIP IF…
- You have never set up Hermes agent — this video assumes an existing Hermes install and skips all onboarding.
- You need enterprise-grade email automation at scale; the n8n workflow shown is a basic proof-of-concept with no error handling or retry logic.
- You are looking for a privacy-first local AI stack — this integration routes queries through Google NotebookLM's cloud and your browser session cookies.
The full version, fast.
Combining the Hermes Telegram agent with Google NotebookLM turns a free 300-source research platform into a portable AI assistant that reads, writes, and acts from anywhere. The mechanism is a two-part skill install: an unofficial NotebookLM skill captures a browser cookie so Hermes can query any notebook programmatically, then a second skill pasted into Hermes grants Telegram-side access to create notebooks, pull infographics, and run audio overviews on demand. Layer n8n's MCP server on top by passing Hermes a bearer token through the terminal, and you can prompt scheduled workflows into existence � like an 8 a.m. email that pulls a daily insight from a chosen notebook � without leaving the chat, at roughly zero query cost.
Members feature
Chat with this breakdown.
Modern Creator members can chat with any breakdown — ask for the hook, quote a framework, find the exact transcript moment. Unlocks at T2: refer 3 friends + add your own API key.
Create a free account →Where the time goes.
00:00 – 00:38
01 · Hook + credibility brief
Capability claim plus outcome promise (light years ahead of competitors). Social proof: built and sold a life-tech startup.
00:38 – 05:11
02 · Why this matters -- 6 use cases
Six illustrated benefit slides: phone access, second brain chat, voice notes, read-to-action, set-and-forget loops, every tool in one chat. Live Hermes OS demo shows YouTube strategy hook principles returned instantly.
05:11 – 07:37
03 · Setup walkthrough
Download free NotebookLM skill, upload to Claude Code or Hermes, authenticate via browser cookie (unofficial Google API), validate connection by listing notebooks.
07:37 – 10:09
04 · Live demo -- query and create notebooks
From Telegram: confirm skill, fetch latest infographic, create brand-new notebook on effective weight loss with 9 sources -- all hands-free from phone.
10:09 – 15:48
05 · n8n MCP extension
Connect n8n new MCP to Hermes. Secure token via bash env var. Build daily-email workflow: 8AM trigger, pull YouTube strategy insight from NotebookLM, send email. Live proof in n8n executions panel.
15:48 – 16:25
06 · CTA + outro
Plug for Hermes OS video (next in series) and Claude Code masterclass in community.
Lines worth screenshotting.
- Connecting NotebookLM to Hermes Agent via a browser-cookie skill lets you query any of your notebooks directly from a Telegram message — no browser required.
- NotebookLM supports over 300 sources per notebook for free, can produce audio overviews, mind maps, and video overviews, and cites every claim — making it the highest-density research tool available at zero cost.
- Hermes can query a NotebookLM notebook and then take action on the answer — drafting a script, saving notes, scheduling reminders, or sending an email — all within the same chat session.
- A cron that queries your NotebookLM at 8am and delivers a podcast overview and infographic to your Telegram is a daily briefing that runs without anyone present.
- Voice-note input to Hermes that creates a new NotebookLM notebook and populates it with relevant sources turns the walk back from the gym into a research session.
- The integration's real power is moving from reading to action — NotebookLM provides the answer, Hermes executes the consequence, in one unbroken Telegram thread.
- Setting up the NotebookLM skill in Hermes takes a few minutes by downloading the skill file and uploading it to Claude Code — the ongoing value massively outweighs the setup cost.
- Hermes operating as a chief of staff with NotebookLM as its knowledge layer means the assistant can answer detailed questions about any topic you have ever researched, from anywhere in the world.
- An n8n MCP integration extends the Hermes-NotebookLM chain to send automated daily emails without the user initiating the send — the system runs the full loop independently.
- The combination of Hermes plus NotebookLM plus n8n MCP is the three-layer stack for autonomous research, synthesis, and action — each layer handles the step the other two cannot do.
- Capturing an idea by voice note in Telegram and having Hermes turn it into a populated NotebookLM notebook immediately removes the gap between inspiration and research infrastructure.
- NotebookLM without an agent is a static analysis tool; Hermes without NotebookLM is an active but shallow agent — the integration makes each one do what it cannot do alone.
A Free 300-Source Research Platform Wired to a Telegram Agent Creates a 24/7 Research Assistant
AI agent workflows
Jack Roberts's walkthrough shows that combining a free multi-source knowledge base with an AI agent accessible from a phone creates a research and automation system that most tools charge subscription rates for — and that n8n MCP extends it into scheduled daily workflows without additional configuration.
02Why this matters -- 6 use cases
- Six use cases: phone access to knowledge, second brain chat, voice notes to action, read-to-action automation, set-and-forget loops, all tools in one chat
- The unifying value is accessibility — a research platform that requires desktop navigation becomes useful on a phone when the agent handles the navigation
03Setup walkthrough
- Download the NotebookLM skill, upload to the agent, authenticate via browser cookie — the unofficial Google API path is the only connection method currently available
- Validate by listing notebooks before building any workflow — confirmation that the connection works is the prerequisite for trusting any automated output
04Live demo -- query and create notebooks
- From Telegram: confirm skill availability, fetch the latest infographic content, create a new notebook with 9 sources — all hands-free from a phone
- The demo proves the workflow works on mobile without desktop access — the phone interface is the primary value proposition
05n8n MCP extension
- Connect n8n MCP to the agent with a secure token stored as a bash environment variable — the secure token approach prevents credential exposure in agent configurations
- Daily email workflow: 8AM trigger → pull NotebookLM insight → send email — three steps in n8n create a recurring information delivery system that runs without manual input
Terms worth knowing.
- NotebookLM
- A free Google research tool that ingests up to 300 sources per notebook and lets you query them, generate audio overviews, mind maps, slide decks, and infographics grounded in those sources.
- Hermes agent
- A personal AI assistant accessible through Telegram that can chat, run skills, and take actions on the user's behalf from any device.
- Skill (AI agent skill)
- A packaged set of instructions and capabilities that extends an AI agent with a new ability, installed by uploading a file and giving the agent a system prompt to follow.
- RAG
- Retrieval-Augmented Generation, a technique where an AI model fetches relevant passages from a source library before answering, so responses stay grounded in real documents rather than guesswork.
- Vector database
- A storage system that indexes text by mathematical meaning so an AI can quickly retrieve the most relevant passages from a large library when answering a question.
- Browser cookie auth
- A workaround that signs into a web service through a normal browser session and reuses the saved cookie to let scripts or agents access the account programmatically when no official API exists.
- Claude Code
- Anthropic's command-line coding agent that runs on your computer, executes shell commands, edits files, and can be extended with skills and MCP integrations.
- Codex
- OpenAI's command-line coding agent, used similarly to Claude Code for running AI-driven development tasks from a terminal.
- Audio overview
- A NotebookLM feature that turns uploaded sources into a two-host podcast-style conversation summarizing the material.
- Mind map
- A branching visual diagram that organizes a topic's concepts and subtopics around a central node, used in NotebookLM to map the structure of a notebook's sources.
- System prompt
- The hidden instruction block that defines an AI agent's role, rules, and capabilities before any user message arrives, shaping how it responds throughout a conversation.
- Persona
- A saved configuration in an AI dashboard bundling a name, description, model choice, and skill set so the agent can be summoned as a specific character or specialist.
- n8n
- An open-source workflow automation platform that connects apps and APIs through visual node-based workflows, similar to Zapier but self-hostable.
- MCP
- Model Context Protocol, an open standard that lets AI agents connect to external tools and services through a uniform interface, so one agent can drive many apps.
- Bearer token
- An access credential passed in an API request header that proves the caller is authorized; whoever holds the token can act as the account owner until it's rotated.
- Access token rotation
- Replacing an existing API token with a freshly generated one, used to revoke leaked credentials or refresh expiring keys.
- Infographic (NotebookLM)
- A visual one-page summary NotebookLM generates from a notebook's sources, laying out key facts, stats, and structure as a shareable graphic.
- Workflow execution
- A single run of an automation workflow, viewable in n8n as a step-by-step log that shows which nodes fired, what data passed through, and where any errors occurred.
- Nondeterministic
- Producing different outputs for the same input on repeated runs; AI language models are nondeterministic by nature, which is why sensitive actions like sending email often get routed through fixed workflows instead.
- Obsidian
- A local-first note-taking app that stores notes as plain Markdown files, often used as an external memory system that AI agents can read from and write to.
Things they pointed at.
05:11linkHermes NotebookLM Skill (free download in description)
07:37productPanfilm (Hermes OS dashboard)
Lines you could clip.
00:06
“I did this and it blew my mind. And it unlocks capabilities that most people on the planet do not even know exist.”
09:40
“The NotebookLM is 100 percent free. We can have a vector database of 300 separate YouTube videos on anything you want and we can query it at zero dollars.”
04:08
“It effectively becomes your chief of staff, your executive assistant has had a incredible knowledge upgrade.”
Word for word.
metaphor
00:00Imagine combining Hermes agent with the world's number one research and intelligence platform, Notebook LM. I did this and it blew my mind.
00:09And it unlocks capabilities that most people on the planet don't even know exist. And in this video, I'll show exactly how to supercharge Hermes agent by powering it with Notebook LM, giving you a twenty four seven AI research agent with new capabilities, saving you time and getting you light years ahead of your competitors.
00:27And if you're new, I'm Jack. I built and sold my life tech startup with a gazillion customers. Now I'm building my own AI startups and I share here the stuff that actually works.
00:35So if you haven't already, grab that beautiful coffee and let's dive straight in. Hermes and Notebook Alam is so fun.
00:41And if you've not heard of Notebook Alam, I'm about to bring you out from the beautiful rock that you've been under. We're gonna cover things like the operating system for memory and Hermes operating system. So many cool freaking things.
00:52But let me go and explain to you why this is so freaking cool. Now, the way that I want you to think about this, because this is gonna unlock new capabilities and it's so epic that you can do this from anywhere in the world. Hermes is your personal AI agent.
01:04Okay? We know this for a fact. We know it's fantastic.
01:06It grows with you. The more that you use Hermes, the better it gets. It's reachable anywhere and it levels up.
01:13It's really freaking cool. It's the best on the market right now. Then we have the wonderful Nelbuk Elam.
01:18As we know very simply, it is the world's number one research and intelligence platform. We can do audio overviews, slide decks, video overviews, mind maps. It can actually collate over 250 different resources on any topic in a planet and you pay precisely $0 for the privilege.
01:35So the idea is can we bring this beautiful knowledge system into Hermes? The answer is, you might have guessed it, yes, we can. Now look, over 300 sources per notebook, $0 to use, and you can host two host podcast audio overviews.
01:47It cites everything, it creates mind maps, it ships artifacts, and you can also get all that stuff directly in your Hermes chat.
01:56It will send it to you in your actual telegram chat. How cool is that? Any podcast you ask for, it is such an overpowered integration.
02:03Now why is this so important? I'm gonna show you, and it's so important to understand why this is key, and then we're gonna actually get this whole thing running up together with some pretty freaking amazing use cases. So first of all, as we can now use NotebookLM from our phone, I can tell you, I have built so much, so many research systems.
02:20There is just something about NotebookLM that just gets levels of research and data and insights that is just so difficult to get when you're standardly searching the Internet. So we can literally now drop a telegram message, Hermes can query any notebook for us, and basically bring back anything that we want to. That's how incredible it is.
02:36And say for example, I'm on the go, I can come over here. This is my Hermes operating system. Of course, you can do this directly in Telegram.
02:42But I might say something like, hey, dude. Could you do me a favor? Just go ahead and check for me my notebook LM on YouTube strategy, and just give me two principles about creating fantastic hooks that are gonna be valuable for me.
02:53Now I have one that has got a insane amount of information on this. Now this agent's gonna work in the background, but effectively, we're on our phone, we're in the gym, we can drop these messages anywhere that we go.
03:03So this is really important. Second of all, we can chat with our second brain. In NotebookLM, if we wanted to, we could effectively build anything.
03:11So you could build brains on whatever things that you want to. For example, I've got one say I did an as an example, right? NotebookLM capability research 50 different notebooks.
03:20Ecommerce, 70 notebooks. There's insane number of things that we can query these insane actual like, literally, like, Alexandra's library materials really easily.
03:29So we have the second brand. Third thing is we can send a voice note and get an answer. So we can now actually grab our phone and yap to notebook alarm.
03:37I think that's really understated just to say, hey, I want to go ahead and create for me a Notebook Alam on the best place to buy Ralph Lauren jumpers or literally, you think of it, we can now build Notebook just by talking to it, which is really awesome. What we can do here is not just read it, but actually going from reading to action.
03:54So note buck alarm answers, Hermes actually can draft scripts, save the notes, schedule reminders, ship the email all from the same chat. So not only can we get the information from note buck alarm, but then within Hermes itself, that can then trigger things like sending emails, which I'm gonna show you in this video. Or whatever the things you wanna do, we're getting the knowledge, and then we're using that knowledge to actually act.
04:13We can set it once, run forever. So literally, I could say every morning 8AM, go ahead and quote me a notebook alarm on something that you think is gonna be valuable based on our conversations in the last twenty four hours.
04:25Then I want you to drop for me the podcast overview and the infographic from that so I can read it first thing at 8AM. That's literally what we can do now with this. That that's where we're at.
04:34It's freaking ridiculous. We've got every tool in one chat, and it can capture your thoughts on the go. So idea hits you in the wild, you can just literally rock up Telegram, give it a thought, and literally that turns into a notebook, and we can grab all those resources straight away.
04:46It effectively becomes your , your executive assistant has had a incredible knowledge upgrade by it basically creating this integration. And then we're gonna come back over here to our Hermes operating system. You can see it's checked our YouTube strategy brain notebook to hook principles and for you.
05:01Make the hook about one painful high stakes problem. Open a knuckle, blah, blah, blah. So you can see, Sentry, it's done it.
05:06It's used my notebook alarm, and it's come back and give me a wonderful answer. So now I know why this is so important. The next thing we're gonna do is actually set this thing up, and the coolest thing, it's so easy to set up.
05:16It takes only a couple of steps, and then you're gonna be rocking and rolling. So the very first thing that we need to do is get the notebook l m skill working for you. So how do we do that?
05:25Well, first thing you do, I'm gonna put a link down below in the description for you for this completely free resource. And by the way, if you really wanna level up and learn Cloud Code, I've got a full master class I'm releasing in the community. It takes you from foundations, building websites, power features, memory systems with Obsidian, building a more Hermes epic stuff apps, all the way down to design systems compliance and turning into dolls.
05:45I'll put a link down below so you can grab this full course. Now, you go over here to grab the free resource, and effectively, you can do is come down and download this Notebook LM skill. You come over here, you click on download.
05:55Now once you've done that, I want you to head over to CoolCode or your model of choice, and effectively come down and just upload that file to it. And then you're simply gonna say, hey there, I'd like you to install the Notebook LM skill, please. And then we're just gonna send that one off.
06:07Now I've already stored it up there, so what it will actually ask you to do is to sign in to Notebook LM. Hey there, could you do me a favor? Can we re authenticate Notebook Alam please?
06:16Now the way that this skill, it's an unofficial skill from Google. So what we need to do is basically open up Notebook Alam in a browser. We sign in and this gets the cookie.
06:25And effectively what happens here then is that your laptop, your computer is connected to Notebook LM, which means that it can anything you can do in Notebook LM, you can now do programmatically with Claude code and or Hermes agent.
06:39If you wanna go deep on the Claude code use case, I'll put a link on screen for you so you can go through and see this. I've had a few versions of this integration. This is the best, slickest, most powerful one and super duper easy.
06:50And effectively now, we can summon that BookLM whenever we like to. Okay. Great.
06:53So once you've done that and you've signed in, it should be working fantastically. So you can just test this works by saying, hey there, do me a favor. Find what are the last three notebooks that I have in NotebookLM.
07:03Okay. Come over and do this. Now bear in mind, there are two ways we use AI.
07:07There's the on our desktop using ClaudeCode or Codex or whatever, and then we have Hermes agent. In the last video, showed you how you connect them together. So you have one universal intelligence system.
07:19It effectively means you get this beautiful network of memory. Everything that we do in Hermes and called code is connected and you get this full overview. I'll put a link on screen so you can just see how you can use together.
07:29It is an absolute cheat code. And as you can see, it's come back and it's literally gone everything. So the next thing we need to do now, now we've validated that we've got the connection between notebook, alarm, and basically Claude code on our computer, we need to connect it to Hermes.
07:42Now to use this, we wanna give it a skill. It's gonna come down to this link down below in the description. And effectively, this will just give a description and a system prompt.
07:49Two ways we can do this. Number one is to copy all the stuff and then head over to our beautiful Hermes agent and basically say, hey, if they turn this into skill, do something else. Personally, I like to use the Panfilm, which lets me basically build out these beautiful characters and I can see physically the different skills that I've got.
08:03So if you are using the dashboard, you'll just create on new persona. You can click on subscribe to this one and and effectively just copy and paste those different bits. Now, the reason why I like to do this for skills that I'm building is I have a nice visual overview of the things that I'm doing.
08:16So I can add a description here, I can come down, grab the system prompt, and then I'm ready to rock and roll. I find it helpful because if I ever wanna change it, I've got a nice visual way of doing it. I am gonna elect to have Ophus 4.7 for the model simply because it's quite good.
08:30And then I'm happy with activating notebook LM skill, click on create scribe, which is fantastic. And then we have describe, which is gonna be on the Book LM skill, which is fantastic. And so to activate this, I'm just gonna copy this prompt here, head back over to Hermes, and just throw this bad boy in here, which is cool.
08:44And so now Hermes has access to the world's best research and intelligence agent. The skill now is how you actually use it to unlock really cool capabilities. Capabilities.
08:52So So first thing I'm do is come down and do this, which is, hey there, dude. I just wanna confirm that you have access to the Notebook LM skill with the scribe. Once it's done that, I'm gonna ask it something else, which is what I'd love you to do for me actually, dude, is go ahead and grab for me the latest infographic that I created on NotebookLM.
09:08And then what I'd like you to do is to create for me a brand new NotebookLM on how to lose weight effectively and bring in all the resources you need to and use any information about me that you may have to make that a reality.
09:22Thank you. And send that off. So now I'm doing two things.
09:25I'm asking three things actually. Number one is confirm you know it. Number two is go ahead and grab me a resource from Netbook I am.
09:31And then thirdly, go ahead and actually create for me a brand new notebook. Now I could be on the go. We could be having coffee with our buddies, right, playing fetch with our dog, and it will go ahead and do all this stuff for us from the laptop wherever we are.
09:43And the reason why this is so cool is because this is superior. There is nothing better. You can search where you want to.
09:50The NotebookLM is 100% free. We can have a vector database of 300 separate YouTube videos on anything you want and we can query it at $0 because we send the question and Notebook LM does all the heavy lifting and the retrieval and the rag and it just spits back the answer.
10:07Meaning that your cost on this is unbelievably low and the answers and quality are insanely powerful. And just like that, it's come back. It took it a few minutes because it was creating and searching.
10:17And look at this, it's grabbed our latest notebook infographic and it's created a brand new notebook, which is awesome. And it's found the actual infographic, which I'm gonna scroll down to show because it's got some different information here. And there you go, guys.
10:28You can actually see this here, and this was something they actually created. So you can see, we can create infographics and notebook alarm for $0 and bring them through. So maybe I wanna learn about, I see you know, maybe one of my buddies was in Rome at the moment.
10:40Maybe he sees a Colosseum and is like, I wanna learn about that. Go create me Notebook LM, do loads of research, and just make me an interesting infographic. You drop that in as a voice note, and then it comes back with this.
10:49You see how crazy this is? Then if I actually head over to NotebookLM, you can see I now have effective weight loss for Jack, evidence based blah blah blah, and it's got nine sources. And I could say go ahead and find 20 sources, 30 sources, it will do that.
11:02Now I can actually query it and ask this notebook whenever I want to. Now we can even take this one step further by integrating n eight n with this. So Cluck Code, we can build a gazillion things.
11:12Hermes Agent, we can build a gazillion things. But there's still a number of use cases where we want granular level control of what's happening, and a perfect example for that is actually sending out emails. So let me show what I mean.
11:24And this is how cool Hermes agent is, is I can literally give it a prompt like this. Hey, n a ten very recently released an MCP. I'm gonna give you the credentials to connect to that n a ten MCP, and I want you to give me a command that I can run-in the terminal that will allow me to give you the full credentialed access to create any scenario I want to check, and I can ask it any questions I want to at any time.
11:47I'm gonna send that one off, and it will do that for us. So within n x 10, we can actually build these whole scenarios. Right?
11:52Where we have an email, we can see the chain, we can see the logic, and actually go back. The coolest feature, I think, is the executions because we want a visual layer sometimes of what we're doing and we can see that in it if we need to. Now if you come over to any 10 dot basically, dot any 10 dot cloud forward slash settings forward slash m c p.
12:10This is the new thing which is cool. So effectively, if we do wanna configure and we do want to grind all the details, we can use Home as agent to do that, which is freaking sick. So on the left hand side, instance level m c p.
12:21Come over, click on enable workflows. So here we're do is come down to connected clients, and you're gonna see there's Lovable, where we did that one previously. But come back to connection details, and then you got the server URL, and you got an access token.
12:32So now what I want you to do is come down and copy this configuration JSON, and then come back over to Hermes. We're gonna paste that in there. I'm gonna come back over and you can see you've got your server URL here.
12:41Come back over, come all the way down and say, and here is my server URL. Okay. And then literally paste in here.
12:47Now I'm gonna show you how you can connect your credentials in a very safe way because when we're using this AI assistance, it's just best practice not to put your API keys in it. Okay?
12:56So I'm gonna down here. I'm gonna say, give me a bash command that I can enter into the terminal with my token that means that you can access it and you have full functionality to do that. Okay?
13:07And then send that one off. As you can see, it's come back and given us all this information. And all you're gonna do is come down and click on copy and just come down to the message and type control v and then this is what we need to enter into the terminal.
13:18So what we need to do is basically replace this with your access token. And you can find that over here by rotating the access token and then just simply copying it and then coming back over. Now we've got that.
13:28All we're gonna do is come down and copy and paste this information here, and then just paste that into the chat window so we can edit it properly. And then we're just gonna replace and add in our token from n eight n. And just so you know where that is, it's literally bearer space, and then you can delete this section here.
13:43Okay? And then you can literally paste in that long number. Great.
13:47And once you've done that, you can do command and t. You can type in terminal. The terminal does appear, and then just copy all this text right here, control c, come over, and just drop that straight into the terminal.
13:56And when that's done, just come down and hit enter. And now Hermes has given us a slightly refined version, is gonna be better. And you can see after entering that in, it says paste n a 10 MCP access token.
14:05So we'll just do that right here inside the terminal. Just like that, it says added n a 10 MCP to Hermes config, which is great. And now we can say, awesome.
14:12Could you go ahead and validate that this is fully added in, please, and that you have access to n a 10? It's gonna come down, ask if we wanna give it permission. We're gonna say that sounds great.
14:19Go ahead. And just like that, it's got all the access. It can search, execute, get search.
14:23If you have clients with any 10, you can now manage it in Hermes. But check this out, if I come down here now, I need to do reload MCP. Let's just do this.
14:30If you didn't know, by way, in Telegram, you have these fantastic little actions you can do, which is fantastic. Yep. Come down here.
14:36We're gonna improve that one. That's awesome. So here's an interesting thing I could do now.
14:38I could present but give it a workflow. So let's say this. Hey, that.
14:41What I would like you to do for me is create a workflow in NA 10 that triggers every morning at 8AM, and it sends me an email to my email address, and that is hard coded into the send file. And then what I'm wanting to do is grab I want you to grab and run, like, one insight from YouTube strategy and send that as, a daily blast in my email.
15:02Create the entire workflow, and then I wanna link to that workflow for me to have a look, and then I wanna actually go ahead and see what this looks like. So you get the idea now. We can build any of these workflows.
15:10And just like that, it has literally gone ahead and done that in a couple of minutes. I'm gonna click on this link to have a quick look at what is going on here, and we'll see what's up. Why is this significant?
15:18Well, let's say that we're managing something here now for a client of subscription. Granted, this is a first it's even got a daily YouTube strategy thing here. This is freaking cool.
15:27I like this. Every morning at 8AM, pick YouTube strategy insight, come down here, and it commends us. So we can understand workflows that break for clients.
15:33If we've got these in situ, very freaking cool. And then if we come down, it's even got the body preview. It shows everything like through the call, and I can even publish that if I want to.
15:40Again, for me, guys, just when it comes to sending emails, just by nature of these models, because they are, as you can understand, like, nondeterministic. Like, they you ask a 100 questions, you get a 100 responses. So I'd like to build out workflows that are you know, I can actually see and understand if I am sending emails out.
15:57And bear in mind, the capabilities now that we have with this Notebook Llab quoting incredible knowledge, using that in any number one of our workflows is ridiculous. I'm gonna be using this literally all the time. It's so insane when you know how to apply it in the right way.
16:11Now using the world's best intelligence and research platform is one thing, but if we don't connect it and have a Hermes operating system, we're leaving too much volume on the table. So the next thing we need to do is get one of those, and we're gonna learn that by watching this video right here.
The Hook
The bait, then the rug-pull.
What happens when a Telegram-based AI agent gets a direct line into a free, 300-source research database? Jack Roberts found out -- and the answer is a 24/7 research-and-action system that runs from your phone while you are at the gym, on the go, or playing fetch with your dog.
Named ideas worth stealing.
03:30model
Research to Action Model
NotebookLM handles knowledge retrieval (free RAG, 300 sources). Hermes handles action execution -- send email, schedule reminder, draft script. Two clean layers: know then do.
Steal forAI stack presentations, landing pages, tool combo pitches
04:17concept
Set It Once, Run Forever
Morning automation: every 8AM Hermes queries a NotebookLM notebook and returns podcast overview and infographic. The integration compounds over time.
Steal forMorning routine content, automation demos, subscription pitch decks
04:35concept
Every Tool In One Chat
Hermes as universal orchestration layer. NotebookLM plus n8n plus Obsidian plus GitHub plus Telegram all accessible from a single chat interface. Zero context-switching.
Steal forAI agent positioning, Claude Code tool demos, productivity stack pitches
How they asked for the click.
15:48next-video
“Now using the world best intelligence and research platform is one thing, but if we do not connect it and have a Hermes operating system, we are leaving too much volume on the table.”
Clean logical bridge CTA -- the n8n demo surfaces the need for a fuller Hermes OS, making the next video feel like a natural necessity rather than a pitch. No subscribe beg. Well executed.
Visual structure at a glance.
