01What Are Agentic Operating Systems and How Do They Work?

Agentic operating systems are systems where you can have all of your agentic harnesses, all of your agents, and all of your systems plugged into one place. Everything is just one click away. This means you can manage your AI team from a single dashboard.

For example, the Hermes Astros system shown in the episode comes up with customized content ideas based on your competitors and keywords. It gives you unique content angles and titles. You can then plug those ideas into your SEO system, video agent system, or notebook.

It also has a powerful memory system so everything learns and stores memories in one place.

02Key Features of an Agentic Operating System

An agentic operating system brings together many useful features. Here are the main ones shown in the episode:

  • Centralized workspace: You have a terminal, a workspace, and a history all in one place.
  • Memory system: All your agents share one brain, so nothing gets lost.
  • Automation builder: You can build new automations based on your system.
  • Command universe: A map that shows every department, agent, service, automation, and system in one place.
  • Image and video studios: Generate beautiful images and videos with tools like GPT Image V2, Minimax, or Grok.
  • Free options: You can use free AI coders and a local coding engine to run without using tokens or costs.

03How to Set Up Your Own Agentic Operating System

Setting up an agentic operating system is not as hard as it sounds. It comes as a quick zip file that your agent sets up for you. Here are the steps shown in the episode:

  1. Get the agentic operating system zip file from the classroom or community.
  2. Install it on your computer or workspace.
  3. Customize it to fit your needs, like changing the user interface or adding new tools.
  4. Set up your agents and connect them to the system.
  5. Use the implementation guide or the one-hour video tutorial to learn the details.

If you want to update later, you do not need to reinstall. Just use the update.md file in the new zip file to apply updates.

04How to Use It: Examples From the Episode

In the episode, Julian showed how members are using agentic operating systems in real life. One member named Mohammed built an SEO machine with a mission control that gives a live view of his SEO and AI workspace. He also created a command universe that maps out every department and agent.

Another example is handling many chats. If you have 20 chats running simultaneously on Claude, you can solve the context problem by having them all work in the same project folder. You can also have your agents log a daily note into an Obsidian folder, so every agent knows what happened in other chats.

Finally, you can fork conversations each time you start a new one, so the new conversation picks up where the old one left off.

05Results and Comparisons From the Episode

The episode did not show formal tests, but it did share real user results. Mohammed's system impressed Julian, and he said the command universe idea was something they might build in. The episode also compared Hermes to Open Core.

Julian said he has not touched Open Core for months because Hermes is so good. Open Core can cost a lot of time in repairing, and it broke after a month or two.

For image generation, the episode compared options. GPT Image V2 is the best generation tool, but Minimax and Grok also work well with Hermes. The key is to use the same API every time for consistent results.

06Tips, Limits, and Mistakes to Avoid

Here are some tips and warnings from the episode:

  • Do not force autonomy: Hermes is not that proactive, but it can run on a schedule. Giving it a task to improve its own automations may go off the rails, so only do that if you are comfortable.
  • Use templated prompts: For better images, use a prompt template that works, and stick to the same image API.
  • Create custom skills: With Hermes, you can create a skill for image generation and update it with feedback each time.
  • Keep chat count low: Limit your daily conversations to around five and fork them to keep context.
  • Do not over-update: If you are happy with your system, you do not need to update it.

07Who Should Use an Agentic Operating System and Why

Anyone who works with AI agents can benefit from an agentic operating system. It is for people who want to run an AI team that works for them day and night. It helps you organize all your agents, tools, and memories in one place.

Even if you are not technical, it is simple to set up because it is just a zip file. The community has over 207 pages of wins from people who are crushing it. If you want to learn more, you can join the AI Profit Boardroom community and get access to the system, trainings, and coaching calls.