← Back to search
NEW Hermes Agent Bot Mode is WILD!!
The AI Firehose · 2026-08-15 · 8 min
Show full episode description
Get the Agent OS & Hermes AI Agent Bot Mode Masterclass 👉 https://www.skool.com/ai-profit-lab-7462/aboutWant to make money and save time with AI? Join here: https://www.skool.com/ai-profit-lab-7462/about
✨ Episode Outline — click any point to jump to it in the episode
Problem solved
Running a team of named AI agents used to mean juggling separate chats and manual copy-paste handoffs;
Hermes Bot Mode puts them in one panel with memory, schedules, and bot-to-bot messaging.
Benefits
- One roster panel with per-bot avatars, memory, and models
- Bot-to-bot messaging with attribution replaces manual copy-paste handoffs
- @-mention any bot mid-chat to delegate work
- Scheduled routines deliver results before you even log on
- Soul (personality) files clone with skills and memory intact
Use cases
- Agent OS roster: Oracle tracks competitors, Astros pulls content ideas from search data, Apollo handles voice replies
- Oracle messages Astros the second a competitor moves on a keyword — no manual handoff
- Muse bot turns rough notes into finished tutorial scripts via @Muse in the main chat
- Cloned research bot watches coaching-call questions and turns them into tutorial ideas
- Morning routine summarizes inbox and checks niche posts daily at noon, results waiting at login
KPIs / results
- One-day public beta launch, free and open source
- 4 coaching calls per week in AI Profit Boardroom
- AI Success Lab community of 75,000 people
- 30-day roadmap for plugging Hermes in alongside other AI tools
Tools / build
- Hermes Bot Mode
- Agent OS (Oracle, Astros, Apollo, Muse bots)
- Hermes built-in scheduler routines
- Soul/personality files per bot
- Agent OS ready-to-install zip
My Agent OS now runs on Hermes AI. Hermes Bot Mode just launched and I plugged it straight into my Agent OS. That's what today's video is about. Technium, the co-founder and lead engineer at News Research, put Bot Mode out today as a one-day public beta. Free, open source, and it changes how you run a team of AI agents. Someone online said it looks a lot like GrokBot. Technium answered with three words. That's the point. So today I'm going to show you my actual Agent OS live on screen. Running with Hermes Bot Mode plugged in, you're going to see the bots, the names, the memory, and how they talk to each other. And by the end, I'll show you exactly how to get this same setup for yourself. Here's what my Agent OS actually is. It's a full folder system I built. Named agents, each one with a job. Oracle tracks my competitors. Astros pulls content ideas from my search data. Apollo handles voice replies. Before Bot Mode, these lived as separate chats I had to click between by hand. Now with Hermes Bot Mode, each one of these agents gets a face, a chat, and a schedule sitting right next to each other in one panel. Oracle can message Astros the second it spots a competitor moving on a keyword. That handoff used to be me copying and pasting between windows myself. Look at the panel on the left. That's the roster. Each bot has its own avatar, its own memory, its own model. My research bot runs one model. My writing bot runs a different one. My inbox bot runs a small, fast model right on my own machine. I picked every single one. That's the piece that makes my Agent OS actually work as a system, not just a pile of separate chats. Every bot remembers its own job. Oracle remembers what it researched. Astros remembers my content style. They don't blur together. Watch this part closely because this is the feature that ties my whole Agent OS together. Bot to bot messaging. Every bot has its own inbox. When one bot sends a message to another, it shows up with attribution so the receiving bot knows exactly who sent it. You can also just type an at sign and a bot's name in any chat. I type at researcher, pull the latest data on this topic. My main bot hands that off. The researcher bot digs in with its own memory, sends the answer back, and my main bot folds it right into the draft. One conversation. Two agents did the work. That's exactly how my Agent OS runs day to day. I don't touch most of it. The bots pass work to each other on their own. Quick pause here because this is the whole reason I built this video. If you want this exact setup, the one you're watching right now on screen, you don't have to build it yourself from nothing. Inside the AI Profit Boardroom, you get the full Agent OS. A ready-to-install zip file with the bots already named, already given jobs, already wired to message each other the way you just saw. You plug it into Hermes and it runs. We walk you through the install step-by-step on four coaching calls every week where you can bring your own setup and get it reviewed live. Daily tutorials show you exactly how to customize each bot for your own business. So it's not just my Agent OS, it becomes yours. Link in the comments and description or go to AIprofitboardroom.com. Now back to the system itself because there's more I want to show you. Every bot in my Agent OS also has its own routines panel. A routine is a task that runs on a schedule without me touching it. So summarize my inbox every morning, check my niche for new posts every day at noon, pull fresh content ideas every Monday. I attach each routine to the bot whose job it is and the results land right in that bot's chat history. So when I open my laptop in the morning, my summary is already sitting there waiting for me. Watch, this is Oracle's chat from this morning. It already ran the routine before I even logged on. Under the hood, these routines run on Hermes' built-in scheduler, which already existed before bot mode. Nothing new was bolted on here. That's actually why I trust this enough to build my Agent OS on top of it. It's not sitting on shaky ground, it's using the same core system Hermes already had. Every bot also has a soul file, which is a personality file. It sets how the bot talks, what it cares about, how it acts. I wrote a custom one for each bot in my Agent OS. Oracle talks differently than Astro's. Astro's talks differently than Apollo. And when you clone a bot, the copy takes the personality file, the skills and the memory with it. So when someone joins the AI Profit Boardroom and grabs the Agent OS zip file, they're not starting from a blank bot. They're starting from a bot that already knows how to do the job because the personality file tells it exactly how to behave. Here's an example of how I'd actually use this for growing a community like the AI Profit boardroom. Say I want a new bot whose only job is watching for questions members ask on coaching calls and turning those into new tutorial ideas. I clone my existing research bot, point the copy at that new job, and it starts out already knowing how our members talk and what they care about. That's the kind of workflow the Agent OS is built for, not abstract examples. Actual jobs that save time running a business. Let me walk you through one more example because this is the one that actually changed how I run my day. Inside my Agent OS, I've got a bot called Muse, whose only job is turning rough notes into finished tutorial scripts. Before bot mode, I had to write the notes myself, then open a totally different chat, paste them in, and hope the context carried over. Now I just type at Muse inside my main chat and hand it whatever notes I've got from the day. Muse pulls from its own memory of how I like scripts written, drafts the whole thing, and hands it back into the same conversation. I never leave the chat I started in. And here's the part that matters for anyone running a community. Muse also remembers which topics land well with our members because that memory builds up call after call, tutorial after tutorial. It's not starting fresh every single time. It gets sharper the longer it runs, the same way a real teammate would. That's the whole idea behind the Agent OS bots that don't just do a task once and forget it. Bots that build a memory of your business over time. So every job gets a little faster and a little smarter than the last one. Let's be honest about where this is still rough because bot mode is only a beta. It's a desktop plugin. So it has to run on the machine running the Hermes desktop app, not on a server by itself. Some early testers installed it on a home server and saw nothing because it belongs on the laptop where the app runs. You also can't put multiple bots in one group chat yet. Right now it's one bot per chat plus messaging and mentions between them. And bot to bot delivery is an instant. When one bot messages another, the receiving bot sees it the next time it runs, not the second it arrives. None of that gets in the way of the Agent OS setup you just watched. It still runs. It still schedules. It still hands work between bots on its own. Here's the bigger point I want you to take from this. The tool matters less than the system. Named agents with jobs, with memory, with schedules, talking to each other. That's the shape of where every AI tool is heading. Just happened to be the one that let me plug my Agent OS in today. If you learn how to run a system like this once, how to define jobs for each bot, how to write a personality file, how to split work between agents, that skill carries over to whatever tool comes next. So here's where this leaves you. You just watched my actual Agent OS running live with Hermes bot mode plugged in underneath it. Named bots, separate memory, separate models, and a roster that hands work between agents without me touching it. If you want that same setup, you don't have to build it from scratch. Inside the AI Profit boardroom, you get the full Agent OS, the exact zip file you just watched me run, ready to install. Plug it into Hermes and your bots are named, scheduled, and talking to each other by the end of the day. You get four coaching calls every week where you can bring your own roster and get it reviewed live. Daily tutorials walk you through customizing each bot for your own business. A 30-day roadmap shows you how to plug Hermes in alongside your other AI tools. And with the member map, you can find people near you already running the Agent OS and compare setups. Link in the comments and description or go to AIprofitboardroom.com. And if you just want the notes from this video, plus 100 more AI use cases like this one for free, join the AI Success Lab. Links in the comments and description. You'll get the write-up from today's video, plus access to a community of 75,000 people building systems just like this one. That's my Agent OS running live on Hermes bot mode. Grab it, plug it in, and watch it work for you the same way.