01What Is Hermes Agent Bot Mode?

Hermes Agent Bot Mode is a free, open source feature from News Research that turns your AI chat into a team of named agents. It launched as a one-day public beta. With it, you can run several AI bots side by side, each with its own personality, memory, and model.

Instead of jumping between separate chats, you see a roster panel on the left. Each bot has its own avatar and chat. They can message each other with @mentions.

This lets you build a system where bots hand off work automatically.

In this episode, Julian Goldie shows his own Agent OS running on Hermes Bot Mode. He has bots like Oracle for tracking competitors and Astros for pulling content ideas. Before Bot Mode, these were separate chats he had to click between by hand.

Now they work together in one panel.

02Key Features of Hermes Agent Bot Mode

Hermes Agent Bot Mode brings several features that make it feel like a real team of workers. Here are the main ones shown in the episode:

  • Named bots with avatars: Each bot has its own name, avatar, and chat window in a roster panel.
  • Separate memory per bot: Each bot remembers its own job and history. Oracle remembers research, while Astros remembers your content style.
  • Different models per bot: You can pick a different AI model for each bot. One runs a research model, another runs a writing model, and a third runs a small fast model on your own machine.
  • Bot-to-bot messaging: Bots have their own inboxes. When one sends a message, the receiver sees attribution about who sent it.
  • @mentions in any chat: Type @ and a bot's name to hand off a task. The bot replies into the same conversation.
  • Routines panel: Schedule tasks for each bot to run automatically, like a morning inbox summary.
  • Personality files: Each bot has a file that sets how it talks and behaves. Cloning a bot carries the file, skills, and memory.

03How to Set Up Hermes Agent Bot Mode

Getting started with Hermes Agent Bot Mode is simple, but you need to follow the right steps. Here is how Julian sets it up:

  1. Download Hermes Bot Mode from the official release. It's a desktop plugin, so install it on the machine where you run the Hermes desktop app.
  2. Open the Hermes desktop app and locate the Bot Mode panel.
  3. Create your first bot by giving it a name and a job description. Julian names his Oracle, Astros, and Muse.
  4. Assign a model to each bot. Pick a different one based on the task. Use a small fast model on your own machine if you want speed.
  5. Write a personality file for each bot. This tells the bot how to talk and what to care about.
  6. Set up routines from the routines panel. Attach each routine to the bot that should run it.
  7. Test bot-to-bot messaging by typing @ and the bot's name in a chat.

If you want Julian's exact setup, he offers the Agent OS zip file inside the AI Profit Boardroom. It already has bots named, with jobs and messaging wired up.

04How to Use Hermes Agent Bot Mode for Your Business

Using Hermes Agent Bot Mode is about letting bots work together. Here are the examples Julian shows in the episode:

First, he uses @mentions to hand off tasks. He types @researcher pull the latest data on this topic. The main bot passes the task to the researcher bot.

The researcher uses its own memory to dig in and sends the answer back. The main bot folds the answer into the draft. One conversation, two agents did the work.

Second, he uses routines for automatic schedules. He attaches a routine to Oracle to check for new posts every day at noon. The results land in Oracle's chat history.

When he opens his laptop in the morning, his inbox summary is already waiting.

Third, he clones a bot for a new job. He clones his research bot and points the copy at a new community. Because the clone carries the personality file, skills, and memory, it starts out already knowing how members talk.

Finally, he uses Muse for turning rough notes into tutorial scripts. He just types @Muse and hands over notes. Muse pulls from its own memory of how scripts should be written and returns a draft in the same chat.

05Results and Comparisons from the Episode

Julian compares Hermes Agent Bot Mode to another tool called Grokbot. Someone online said it looks a lot like Grokbot. Technium, the co-founder and lead engineer at News Research, answered with three words: That's the point.

Julian shows real results from his morning Oracle routine. It already ran the routine before he logged on. That means the summary was sitting there waiting for him.

This shows the scheduler works as promised.

He also explains that routines run on Hermes's built-in scheduler, which existed before Bot Mode. Nothing new was bolted on. This makes him trust the system more because it's not sitting on shaky ground.

The episode doesn't show benchmark numbers or side-by-side speed tests. It focuses on how the system changes his workflow.

06Tips, Limits, and Mistakes to Avoid with Hermes Agent Bot Mode

Hermes Agent Bot Mode is still in beta, so there are limits. Here is what Julian says to watch out for:

  • It's a desktop plugin: It must run on the machine with 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.
  • One bot per chat: You cannot put multiple bots in one group chat yet. Right now, it's one bot per chat, plus messaging and mentions between them.
  • Delivery is instant: When one bot messages another, the receiving bot sees it the next time it runs, not the second it arrives. Be patient with handoffs.
  • Write good personality files: The personality file tells the bot how to behave. Without it, you start from a blank bot and lose the system's power.
  • Learn to split work: The tool matters less than the system. Define jobs for each bot and write clear routines.

Despite these limits, the agent OS setup still runs, schedules, and hands work between bots on its own.

07Who Should Use Hermes Agent Bot Mode and Why It Matters

Hermes Agent Bot Mode is for anyone who runs a business or community and wants to automate repetitive tasks. It is especially useful for content creators, community managers, and small business owners who juggle multiple AI chats.

The bigger point Julian makes is that the tool matters less than the system. Named agents with jobs, memory, and schedules talking to each other is where every AI tool is heading. Hermes just happens to be the one that lets him plug his agent OS in today.

If you learn how to run a system like this once, you can carry that skill to whatever tool comes next. You learn how to define jobs, write personality files, and split work between agents. That's a skill that will serve you beyond Hermes.

Julian offers the full Agent OS inside the AI Profit Boardroom, a ready-to-install zip file. You also get four coaching calls weekly, daily tutorials, a 30-day roadmap, and a member map to connect with others.