--- title: 'Use Hermes Agent Like a Pro With These Tricks' source: 'https://youtube.com/watch?v=KSNQ5mwMasM' video_id: 'KSNQ5mwMasM' date: 2026-07-30 duration_sec: 484 --- # Use Hermes Agent Like a Pro With These Tricks > Source: [Use Hermes Agent Like a Pro With These Tricks](https://youtube.com/watch?v=KSNQ5mwMasM) ## Summary The video presents Agent OS as a dashboard for running the Hermes AI agent, promising a 10x power boost over terminal usage. It highlights new updates including a 78% memory reduction and free model integrations, while demonstrating tools like Oracle, Astros, and Apollo for content creation and competitor analysis, ultimately pitching the AI Profit Boardroom membership. ### Key Points - **Terminal vs Agent OS** [00:02] — Running Hermes in a terminal is slow and flat; using Agent OS dashboard turns it into a full operation with shared memory. - **Hermes Cloud Setup** [00:52] — Set up Hermes Cloud in about 60 seconds via a few clicks to run an agent 24/7 in the cloud, automating content for the AI Profit Boardroom. - **Memory Reduction Update** [01:21] — New update shrinks chat session memory by up to 78% (60% average), making sessions load faster and history search quicker. - **Free Model Integration** [01:46] — Hermes now allows free models like Laguna S2.1 and HY3 via the portal; HY3 excels in design and games, Laguna for hard thinking. - **Oracle Tool** [03:08] — Oracle pulls daily trending AI news, categorizes it, and provides original links, helping create content for the AI Profit Boardroom. - **Astros Tool** [03:48] — Astros monitors competitors by analyzing their posts and keywords, scoring winning ideas to create fresh content. - **Apollo Voice Agent** [04:16] — Apollo is a voice agent that listens, talks back, and can build apps, dashboards, and games by voice while transcribing everything. - **Video Agent** [04:41] — Built-in video agent researches trending topics, writes scripts, and produces fully edited animated videos with voiceover autonomously. - **Outreach Tool** [05:09] — Outreach tool finds target audience, pulls details, and drafts personal messages for approval, enabling human-like outreach. - **Agent Brainstorming Room** [05:36] — Multiple agents brainstorm together, building on each other's ideas, then pipeline to approval and building for content generation. - **Stacking Tools** [06:02] — Pro tip: stack tools like Astros, Oracle, and video agent in one chain, then let Hermes Cloud run overnight for automated output. - **Shared Memory** [06:18] — All tools feed into a shared Obsidian vault memory (shown as a galaxy), enabling every agent to know everything simultaneously. ### Conclusion Running Hermes inside Agent OS with shared memory and stacked tools creates a powerful automation system that works while you sleep, far exceeding terminal-only usage. ## Transcript tricks, still running your AI agent in a terminal? Stop. That's the slow way, the hard way. The pros don't touch the terminal. They run Hermes inside one dashboard called Agent OS. Same agent, 10 times the power. Let me show you the tricks most people miss. I'm the digital avatar of Julian Goldie, and I help in their real work. Today, I'm showing my Agent OS. Plus, a brand new update that makes the whole thing faster. By the end, you'll see why one agent in the right place beats 10 agents scattered almost nobody switches on. Here's what most people get wrong. They install Hermes, open a terminal, type a command, and wait. It works, but it's flat. On its own, Hermes is just an agent. But when I drop it inside my Agent OS, it turns into a full operation. Agent OS is the dashboard that holds Hermes and all memory, so everything works together. That's the whole trick. So, let me show first tool I lean on is Hermes Cloud. This lets me run an agent all day and cloud. I set it up in about 60 seconds. A couple of clicks, sign in, and it's employee, and that's exactly what it feels like. I use Hermes Cloud inside my the clock, quietly building content for the AI Profit Boardroom while I'm asleep or busy. Every morning, I've got more done for the AI Profit Boardroom than do that from a terminal. Now, here's the update that makes all of this faster. Hermes just got a lot lighter. Every chat session gets stored in a little it a lot, that thing balloons and slows down. The new update shrinks it by up to 78% and about 60% on average. Your sessions load faster, history search is quicker, and it takes up way less room on the machine your agent lives on. And if it's under a gig, it cleans itself up Now, let's talk brains. Hermes now gives you free models you can plug straight models inside it, and some cost nothing. Two are worth knowing, Laguna S2.1 and HY3. Both are free right now through the portal. I set them up in the the section inside Agent OS, open the model area, pick the news portal, and choose a free lighter version of Hermes, I scroll down in that same manage section and hit headaches. Now, which free model is actually best? We tested it properly Laguna came out kind of average, especially running locally. HY3 was the standout. It makes clean designs. It even put together nice little 3D games. board. So, for agent work and building things, HY3 is the one I'd reach for. S2.1 is stronger on the really hard thinking tasks. So, it's handy to have both. And I'll be honest, these free models aren't quite as sharp as top paid ones like Kimi K3 or Fable 5, but they're free, fast, and in the cloud what I do. Now, if you're watching this thinking you'd love to run Hermes inside don't want to piece it all together yourself, that's what the AI Profit Boardroom, I've put the full Agent OS setup as a zip file you can install with Hermes already wired in and every tool I'm showing you today ready to go. you step-by-step from opening it the first time to running your own agent in where you share your screen, show me your setup, and get unstuck on the spot. models and the cloud setup. So, you're never guessing. If you want my exact scratch, the AI Profit Boardroom is where it lives. Okay, back to the tricks. Let me show you the tools that make Agent OS feel alive. First up is Oracle. Every 24 hours Oracle pulls in the latest trending AI news for me. It's login. So, it finds what's breaking right now, not old stuff. A normal agent news by category, gives me the original link, and here's the clever bit. It story for my own content. I use Oracle every morning to grab the freshest AI that pull the right people toward the AI Profit Boardroom. It can even draft the post or push it straight to a WordPress site. So, the AI Profit Boardroom stays in front of the exact people who care about this stuff without me starting from a blank page. Right next to that is Astros, the competitor radar. I drop a Astros and it studies what they're posting. It reads their keywords, checks what's landing, and scores it. So, the best ideas float to the top. It finds off. I use Astros to spot which AI topics are trending, then I turn the strongest ones into fresh content ideas for the AI Profit Boardroom. Instead of what people want before I start. That's a pro move most people never think to Apollo is a voice agent. I literally briefing, put it in focus mode, and it listens and talks back. It keeps a transcript of everything and it can build things while we chat. I've asked page for the AI Profit Boardroom just by talking out loud. It can build little apps, dashboards, even games, all by voice. I can give it a wake word so it in the background while I do other things. Building something for the AI the keyboard still feels a bit like magic. Now, this next one is big. Agent OS has a video agent built in. I point it at a trending topic and it does the whole thing. It researches the topic, finds the latest news on it, writes the script, and then builds a fully edited, animated video with a voiceover, all in the background. I use this to turn one trending AI topic into a short video the AI Profit Boardroom. It pulls a few tools into one workflow, so I just give a finished video out, and more of the right people learning about the AI outreach tool inside Agent OS. I describe the exact kind of person who'd Boardroom and it goes and finds them, pulls their details, and drafts a personal message. Nothing sends until I approve it, so it always stays real and human. It's a clean way to reach people who'd actually want to learn with us in out of ideas, I don't have to think hard. Agent OS has a room where the agents brainstorm together. One agent throws out an idea, another builds on little team. When something good comes out, it gets saved into a pipeline. From I approve it or delete it, then into building, then it ships. I use this to the AI Profit Boardroom, letting the agents pitch ideas while I just pick the winners. Here's a pro tip most people skip. I don't run these tools one at a time, I stack them. Astros finds the trending topic, Oracle checks the turns that into a finished clip all in one chain. Then I let Hermes Cloud run the time I wake up. That's the real trick with Agent OS. On its own, each one memory, they get through a full day of work while I sleep. That's what like. And here's what ties the whole thing together. All of this feeds into one shared memory. Every tool logs what it learns into an Obsidian vault that Agent OS shows as a kind of galaxy. It's a second brain. When Astros finds a trend or Oracle grabs the news, every other agent instantly knows about it, They're all reading from the same brain at the same time. That's why running Hermes inside Agent OS is 10 times more powerful than typing one command into a terminal. In the terminal, the agent knows nothing about your world. Inside Agent OS, it knows everything. So let me bring it home. Hermes on its own is good, but Hermes inside Agent OS running in the cloud using a free brain, pulling live news, watching competitors, building videos, talking to me by voice, and remembering everything in one shared memory, that's a full operation. That's and running one like a pro. And it's lighter and faster now than it's ever a free model in about 5 minutes. If you want the full process, SOPs, and 100 plus AI use cases like this one, join the AI Success Lab. Link's in the comments and description. You'll get all the video notes from there, plus access to our community of 85,000 members who are crushing it with AI. And if you're about to go try this yourself, here's my honest tip. The setup is easy, but the is wired together and sharing one memory, and that's the part people get stuck on. That's exactly why I built the AI Profit Boardroom. Inside, you get the full Agent OS zip ready to install, Hermes pre-wired, and the 30-day roadmap that takes you from your first click to a cloud agent that runs on its own. You models, the cloud setup, and every tool I showed you today. And when you hit a coaching calls, share your screen, and get it sorted with over 4,000 members building right alongside you. Come join us at aiprofitboardroom.com.