1:01:12Sabrina Ramonov 🍄 · Tutorial
Claude Cowork 60 Minute MASTERCLASS
A complete zero-to-hero tutorial on Claude Desktop's agentic mode: five real use cases, three core primitives, and honest caveats about where it falls short.
April 9thThe argument in one line.
You can monitor, analyze, and post to X without paying for any infrastructure or social listening software by chaining a free cloud AI workspace, an open-source agent, and Brave Search into a single Telegram bot.
Read if. Skip if.
READ IF YOU ARE…
- You want to track trends, competitors, or product launches on X without paying for enterprise social listening tools.
- You are new to AI agents and want a guided build that uses free-tier tools for the core functionality.
- You already have a Telegram account and want on-demand research briefings delivered directly to your chat.
- You have been putting off building an agent because you assumed you needed a dedicated machine or expensive API access.
SKIP IF…
- You need real-time streaming data from X — this workflow scrapes public search results, not the firehose.
- You are looking for a production-grade solution with uptime guarantees and error recovery.
- You have already built agentic workflows with n8n or similar tools — this is introductory-level content.
The full version, fast.
The video builds an X research and auto-posting bot that lives entirely inside Telegram. Zo Computer provides the cloud machine so nothing runs on your laptop; Hermes acts as the agent layer interpreting commands and calling tools; Brave Search with a site:x.com operator replaces the paid X API for the read side of the workflow. Send a topic to the Telegram bot and within minutes you receive a structured briefing covering themes, sentiment, key accounts, and notable posts. A second optional prompt adds OAuth-based X posting for roughly $0.015 per tweet.
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 →Where the time goes.
00:00 – 01:59
01 · Hook & Problem Setup
Introduces the OpenClaw-style agent concept, explains why a dedicated machine is no longer necessary, and previews the stack: Zo Computer + Hermes + Telegram.
01:59 – 02:46
02 · System Demo
Live walkthrough of the finished bot: send a topic via Telegram, receive a structured research report within minutes.
02:46 – 04:55
03 · Setting Up Hermes in Zo Computer
Download Zo Computer, install Hermes via terminal command, select AI provider (OpenRouter free tier), configure two key settings: AI model and Telegram integration.
04:55 – 06:27
04 · Creating the Telegram Bot
Use BotFather to generate a bot token; use RawDataBot to retrieve your Telegram user ID; paste both into the Hermes setup wizard to complete the connection.
06:27 – 10:53
05 · X Research and Post Automation
Paste Prompt 1 (free research layer via Brave Search, no X API key required). Test with a live query. Optionally paste Prompt 2 for X posting via OAuth 2.0 at ~$0.015/tweet.
10:53 – 12:09
06 · Final Thoughts & Conclusion
Stack recap: Telegram (interface), Brave Search (discovery), Hermes (orchestration), AI (analysis). Closing CTA to join the Any No Code community.
Lines worth screenshotting.
- The X API is not required to monitor public X conversations — Brave Search with a site:x.com operator retrieves public posts for free.
- Zo Computer is a cloud workspace that runs AI agents without requiring a dedicated machine, eliminating the Mac Mini or VPS setup problem.
- The entire research layer in this workflow costs zero dollars: no API key, no subscription, no social listening platform.
- X posting via the official API costs approximately $0.015 per tweet — negligible unless publishing at massive scale.
- 80% of the value in this system comes from the free research layer alone; the posting automation is optional.
- Hermes translates a single Telegram message into a multi-step pipeline: search, parse, analyze, report — no user involvement required between trigger and output.
- The bot token from BotFather is effectively an API key for your Telegram bot and must be kept secret.
- OpenRouter provides a free AI model tier that serves as the reasoning layer for Hermes, making the entire base stack free to run.
- Structuring AI output as a briefing — themes, sentiment, key accounts, notable posts — rather than a list of links is what makes the output actually useful for decisions.
- OAuth tokens for X posting rotate on every use and must be persisted to a JSON file, not environment variables, or the auth loop breaks on restart.
Public X data is free — the barrier was always the workflow.
WHAT TO LEARN
Brave Search with a site:x.com operator returns public X posts without any API key — and chaining it to an AI agent turns raw search results into a formatted intelligence briefing.
01Hook & Problem Setup
- Cloud AI workspaces eliminate the need to buy and maintain a dedicated machine for running persistent agents.
02System Demo
- Seeing the finished output before the build tutorial helps you evaluate whether the effort is worth it for your use case.
03Setting Up Hermes in Zo Computer
- OpenRouter provides a free AI model tier that serves as the reasoning layer — the base stack costs nothing to run.
- The two settings that matter in Hermes setup are AI provider/model selection and the Telegram bot connection; everything else is default.
04Creating the Telegram Bot
- BotFather issues a bot token; RawDataBot returns your user ID — both are needed before Hermes can connect to Telegram.
- The bot token is the API key for your Telegram interface and must be kept secret like any credential.
05X Research and Post Automation
- BraveSearch with a site:x.com operator retrieves public X posts for free — no X API key, no developer account, no subscription required for the research layer.
- Structuring AI output as a briefing rather than a list of links is what makes the output actually usable for decisions.
- The free research layer delivers roughly 80% of practical value; the paid posting layer is optional and can be added later.
- OAuth tokens for X posting rotate on every use and must be persisted to a JSON file rather than environment variables, or the auth loop breaks on restart.
06Final Thoughts & Conclusion
- The stack is intentionally minimal: Telegram, Brave Search, Hermes, AI — four components, no proprietary platform lock-in.
Terms worth knowing.
- Zo Computer
- A cloud AI workspace that lets you run agents and tools in the cloud without maintaining your own server or keeping a laptop online.
- Hermes
- An open-source AI agent that can use tools, follow multi-step instructions, and connect external services — acts as the orchestration layer between Telegram commands and tool calls.
- Brave Search scraper
- A technique that queries brave.com with a site:x.com operator to surface public X posts without using the X API or any paid credentials.
- OAuth 2.0 with PKCE
- The authentication flow required by the X v2 API for posting. Generates short-lived access tokens and long-lived refresh tokens that rotate automatically on each use.
- BotFather
- Telegram's official bot for creating and managing other bots. It issues a bot token — the API key that links your Hermes agent to your Telegram channel.
- OpenRouter
- An API aggregator that routes requests to multiple AI providers. Offers a free-tier model that Hermes can use as its reasoning engine at no cost.
Things they pointed at.
00:37toolHermes Agent
05:05toolBotFather (Telegram)
05:45toolRawDataBot (Telegram)
Lines you could clip.
00:49
“No more X API needed. No expensive social listening software. No communicator infrastructure anymore.”
07:28
“We would recommend that everyone start with the research layer first because, honestly, that's where most of the value comes from.”
09:25
“Imagine if this were your morning workflow — you could spend thirty seconds reading this report and immediately understand what's happening across the AI space without opening X once.”
11:25
“The data was always available. The barrier to understanding it just became dramatically smaller.”
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.
metaphoranalogystory
00:00What if I told you that you can build an OpenClaw style social listening agent without buying and setting up a separate computer? With OpenClaw, many people start by buying a clean machine such as a Mac Mini, installing the right tools, configuring the environment, and keeping the computer online.
00:16And now with Zoho computer, you can launch your own cloud computer in one single click, and there is even a free plan so you can start testing it without paying anything. So in this video, we're going to use Zoho computer and Hermes agents to research conversations on x, find the main themes, analyze sentiments, identify influential accounts, and pull out the most important post.
00:39And before I start, let me quickly introduce what Zoho computer is. So you can think of Zoho computer as a personal cloud computer for AI agents. Like, it can run tools, handle tasks, and keep your work in the cloud without relying on your own laptop.
00:54So with this research system setup, no more xAPI is needed. No expensive social listening software. No communicator infrastructure anymore.
01:02Just Zoho computer, Telegram, Hermes, public web search, and AI analysis working altogether. And by the end of this video, we are going to show you how to publish content back to X automatically as well. So let's get into it.
01:16So the system that we are going to build today is an AI powered X research assistant built with Zoho computer and Hermes agent, and it lives entirely inside Telegram. And normally, researching a trending topic on x means, like, opening the app, searching manually, scrolling through post, reading reply chains, and trying to figure out what actually matters.
01:35And with this workflow, you can simply send a question to a Telegram bot, and Zoho computer runs the agent in the cloud, premise searches public x conversations, gathers the most relevant posts, analyze the discussion, and then sends a structured research briefing back to you directly inside Telegram. Basically, can use it to research a product launch, track a competitor, understand a viral debate, or monitor specific accounts.
01:59And before we build anything, let me show you what the finished system actually does. So here we are inside Hermes agent, and we are running everything inside Zoho computer. You can think of it as a cloud workspace where we can run agents, connect tools, and automate workflows without, like, setting up your own infrastructure.
02:16Let's say I can send a message to this Telegram bot and ask it to research pretty much any topic, and then Hermes will go and find the most relevant post, analyze the conversation, and return a complete report directly inside to Telegram. Once the research comes back, we can ask Hermes to turn those insights into content, and then it can summarize the findings, draft a post, and publish it to x automatically.
02:40Everything happens from a single Telegram chat. Now you have seen the workflow. Right?
02:44Let me show you how you can build it as well. Alright. The entire workflow runs using Hermes agent, so that is the first thing we need to settle.
02:52And instead of managing our own infrastructure, we are going to run it inside Zelle computer and connect the tools from there. So if you do not have it installed yet, you can head over to Zelle computer, download the installer, and run through the setup process.
03:07And once it's installed, you will be asked to create an account. Just go ahead and do that. And when you are finished, you should end up inside a workspace that looks like this.
03:17Now before we install Hermes, let me quickly explain what it actually is. So you can think of Hermes as an AI agent that can use tools, follow instructions, and automate workflows on your behalf.
03:29And instead of simply answering questions like a chatbot, Hermes actually can perform actions, connect services together, and execute multistep tasks. And in our case, we are going to use it as the engine behind our X Research and Content Automation workflow.
03:45So with that out of the way, let's grab the installation command and then jump back into Zoho computer. And next, click on the terminal and simply paste the command and hit enter. And from here, you can see that Hermes will launch its setup wizard and guide you through the rest of the installation process.
04:04And most of the options are pretty straightforward, but there are two settings worth paying attention to. So the first one is your AI provider and model selection.
04:13And for this tutorial, we're using OpenRouter together with our alpha because it's completely free and works great for just getting started. And if you already have access to other providers like OpenAI and Thorpec or Gemini, you can simply use those as well. And the workflow we are building today will work the same way regardless of which model you use.
04:34And the second important step is the Telegram integration. And basically, Hermes needs a Telegram bot so it knows where to receive messages and where to send responses. And we are going to create that bot next and then come back here to finish the connection.
04:50So for now, you can just leave the setup wizard opened. Let's create the Telegram bot first. So once we connect Telegram to Hermes, every message we send can trigger actions inside our workflow.
05:01And now let's just connect to Telegram. First, you open Telegram and search for add bot father. This is Telegram's official bot for creating and managing bots.
05:09And once you're inside, just send the command slash start, and you'll see all the available options. And then type or choose slash new bot to create a new bot here.
05:20And from there, bot father will walk you through the setup. You can just choose a name, pick a username, and once everything is finished, Telegram will give you a bot token.
05:29And this token is essentially the API key for your bot, so make sure you keep it somewhere safe because we're going to need it in a moment. Next, we need to get our Telegram user ID, and to do that, you can search for raw data bot and send a command slash start again, and then it will immediately return your user information, including your user ID.
05:51Just copy that as well because we are going to need that for the Hermes setup. And now let's jump back to Zoho computer. And if your Hermes session restarted while we were doing this, don't worry about this first.
06:03Just open the terminal, type Hermes setup, and then continue through the wizard until you get back to the telegram configuration step. And at this point, all we need to do is just to paste in our bot token, enter our Telegram user ID, and complete the setup.
06:18And our Telegram bot is now connected to Hermes and ready to receive commands. Now let's actually teach it how to research on eggs and automate posting. In order to make it easy, we have had two prompts so that you can simply copy and paste into Hermes.
06:33So the first prompt sets up the eggs research layer, and this is completely free. And what makes this interesting is that we are not using the X API tool. And instead, Hermes uses Brave Search to discover public posts on X, gather the relevant information, and then use AI to analyze and summarize the conversation.
06:52And that means no API keys needed, no subscriptions, no paid social listening tools, and no ongoing cost. And the second prompt is that we add XPost automation using XOfficial API and OAuth two point o credentials.
07:07And unlike the research workflow, this part actually does require x developer access because Hermes is actually publishing content to your account through the official API. And the good news is that the cost is extremely low, around 0.015 US dollars per tweet.
07:25So for most people, it's practically negligible. And now for us, we would recommend that everyone start with the research layer first because, honestly, that's where most of the value comes from. And once this is working, you'll be able to ask questions about any topic, any trend, or even a specific x account.
07:42And Hermes will go out to find relevant posts, analyze the conversation, and return a complete research report directly in Telegram. No xAPI, no subscriptions, no social listening platform.
07:54Just public search and AI analysis working together. Now let's start with prompt one. We are going to paste it into Hermes right here and let it run.
08:03And what this prompt is doing is teaching Hermes how to research conversations on eggs from end to end. First, it searches brave search for publicly available posts related to your topic, and then it collects the most relevant results, identifies the important accounts involved in the discussion, analyzes sentiment, and looks for recurring themes across multiple posts.
08:25And finally, it uses AI to turn all of that information into a structured report that's actually useful to read. So instead of getting a list of links, you are actually getting a briefing. And because everything is built on public search results, the entire research workflow costs absolutely nothing to run.
08:42Once the prompt finishes, we can immediately test it, and let's jump into Tellurium and ask the bot something practical. So we are going to ask it for the latest AI news and trends. And within a few seconds, Hermes will start researching the topic, gathering relevant posts from eggs, identifying the major conversations, and looking for the accounts driving those discussions.
09:04And after a couple of minutes, here comes the report. And right away, we can see the biggest topics people are talking about, like which AI products are getting attention, what the overall sentiment looks like, and the specific post that Hermes used to build its analysis. And what's nice about this is that we are not getting a list of search results.
09:22We are actually getting an actual breathing. And imagine if this were your morning workflow, you could spend thirty seconds reading this report and immediately understand what's happening across the AI space without opening x ones. And now if that's all you want, you can actually stop here because you'll probably get about 80% of the value from this entire project with just this free research workflow alone.
09:46But, of course, if you want to take things a step further, we can continue and add posting automation. And for that, we are gonna use the second prompt here. And before you can do so, you will need the x developer credentials.
09:58That means creating a developer application, generating your OAuth credentials, and grabbing your OAuth two point o tokens.
10:05And the setup is pretty straightforward, and prompt two will handle all of that for you. And you just need to get your X developer tokens ready.
10:13And, again, there is a small cost for posting through the X API. And for most people, it's going to be just a fraction of a cent to a few cents per post depending on usage, of course. So unless you are publishing at a massive scale, the cost is pretty negligible.
10:29And once your credentials are ready, just paste prompt two into Hermes and let it configure the posting workflow. And now Hermes can do both sides of the process.
10:39First, research comes in through Telegram, and then we get the insights analyzed, and then content will get generated and posted automatically to x.
10:49And then that right there is the entire system. Let's wrap everything up. So here is what this system actually delivers.
10:56A single Telegram message can actually trigger a complete research workflow. Hermes searches public x content, analyzes Dalsons of relevant posts, summarizes major themes, and delivers a report directly back to your chat.
11:10And the stack itself is pretty simple. First, Telegram for the interface, BraveSearch for discovery, Hermes for orchestration and reasoning, and AS memorization for turning raw information into useful intelligence.
11:25And that is really what changed. The data was always available. Right?
11:29. Try this out yourself so that you can now monitor industries, competitors, product launches, breaking news, or even create a conversations without spending hours scrolling through x.
11:45And if there are other Hermes agent builds or AI automation workflows you want us to cover, let us know in the comment section. And, of course, if you want more in-depth tutorial like this and to actually make more money with AI, feel free to join our Any No Code community. You can find the links in the description below as well.
12:01And as always, if you found this video helpful, drop a thumbs up and subscribe the channel for more video like this in the future. I'll see you in our next one.
The Hook
The bait, then the rug-pull.
The premise is deceptively simple: you want to know what people are saying on X about a topic, without paying for the X API, without subscribing to a social listening platform, and without keeping a dedicated computer online. This video makes the case that all three problems are already solved — and shows you how to wire the pieces together in about twelve minutes.
Named ideas worth stealing.
06:27concept
X Research Layer — Prompt 1
A paste-in Hermes prompt configuring a BraveSearch-based X research pipeline. No paid API keys. User sends topic via Telegram, Brave searches site:x.com, parser extracts URLs/handles/text, AI summarizes, formatted report delivered to Telegram.
Steal forany topic-monitoring or competitor-tracking use case
08:50concept
X Posting Layer — Prompt 2
A second paste-in prompt adding OAuth 2.0 X posting to the existing research workflow. Requires X developer credentials. Tokens rotate on every use. Cost ~$0.015/tweet.
Steal forclosing the loop from research to publishing without switching tools
09:25concept
80/20 Research Stack
The free research layer alone delivers roughly 80% of the system value. The paid posting layer is an optional add-on. Start with the free layer, add posting only when needed.
Steal forany tiered build where you want to de-risk commitment before paying for API access
How they asked for the click.
VERBAL ASK
11:57product
“if you want more in-depth tutorial like this and to actually make more money with AI, feel free to join our Any No Code community”
Soft sell at the very end after all content is delivered. Community link in description. Low friction, appropriate placement.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
AFFILIATECommission earned if you click.
OTHER LINKSAlso linked in the description.
Visual structure at a glance.
