01What Agent OS Is and How It Works

Agent OS is a system that links all your AI agents together with a memory system so they work inside one space. It connects agents like Hermes, Claude, or others so they can share context and build on each other's work.

In this episode, Julian Goldie explains how agent OS helps you automate and build anything. The core idea is that your agents stop working in silos and start collaborating. They use a shared memory, which makes them smarter over time because they remember everything you have done.

Think of it as a central hub. Instead of juggling multiple AI tools separately, you have one system where all agents pull from the same knowledge. This makes automation smoother and more powerful.

02Free Model Options for Hermes Agent

Hermes Agent used to have a free option with Qwen 3.6, but that beta trial period has ended. However, there are other free ways to keep using Hermes without paying.

One option is Omniroot, an open source project that gives you free models across 90 different APIs. Omniroot automatically selects models and has built-in fallbacks, so if one free API gets rate limited or expires, it switches to another one.

Other free model choices include:

  • Step 3.7 Flash
  • Poolside Laguna XS 2.1
  • Ling 3.0
  • Laguna S 2.1
  • Tencent HY 3.3

You can also use your existing subscriptions for Kimi, Minimax, or DeepSeek if you already pay for them. That way you do not need a separate agent subscription.

03How to Set Up Hermes with Free Models

Setting up Hermes Agent with free models takes just a few steps inside the agent OS dashboard. Julian walks through the process in the episode.

  1. Open your agent OS dashboard.
  2. Switch over to Hermes in the main view.
  3. Go to the manage section in your dashboard.
  4. Find the model section and select a new model.
  5. Choose Omniroot or one of the other free models from the list.
  6. If you use a subscription like Kimi or Minimax, select that model instead.

Once you change the model, Hermes will start using your chosen option. If you get stuck, the manage section shows all available free models, so you can always switch to another one.

04Set Up Obsidian as Your AI Memory System

Obsidian is Julian's memory system for AI agents. It stores all context and memories in linked folders and files, so your agents understand everything about you and your work.

To start, make sure Obsidian is set up on your computer. Then ask any AI agent to export its memories and organize them into folders and files inside Obsidian. The agent will create a full setup with everything neatly arranged.

Julian calls this a memory galaxy. You can visualize how agents understand context from the past few weeks. Clicking on a link, like one called LLM Wiki, explains a concept in detail.

Once the memory is in Obsidian, you tell each AI agent to connect to Obsidian locally. Any agent can use it, even ones outside agent OS. This works with Claude, Open Core, Buzz, or others.

05Automate SEO with Agent OS Step by Step

Agent OS can handle SEO for your websites. Julian shows two ways to set this up, depending on how much time you want to spend.

The simple version uses a 12-page SEO skill that you give to your AI agents. They read it and start creating content for your site.

The more powerful version uses the agent OS system:

  1. Get the agent OS SEO zip file inside the AI Profit Boardroom.
  2. Attach the zip file to your cloud agent.
  3. Say, set this up for me and configure cloud so it works in the chat.
  4. Log in when asked and offer the URL.
  5. Plug in your Netlify access token or WordPress login details with an API password.

After that, agents can log into your site and start publishing SEO content. Julian demonstrates a fresh setup done in one morning.

06Results and Trajectory of SEO Automation

Julian shares results from using agent OS for SEO. His websites are on a good trajectory, growing nicely over time.

The system plugs in a keyword and a case study, then deploys content to the website. This works well for multiple sites at once.

He also mentions building an FAQ section for the agent OS setup. It includes examples for configuring CLIs, setting up Obsidian, and using VBS. A one-hour video tutorial goes even deeper for those who want full detail.

07Custom GPTs for Client Workflows

Julian answers questions about building and sharing custom GPTs for clients. His team does this all the time, like creating workflows that come up with new video ideas.

He does not recommend self-hosting an open source model for a 15-person team. Instead, his team shares one company ChatGPT account.

For sharing custom GPTs, you just share the link. You never need to share credentials. One or two account managers create all custom GPTs for clients, which keeps things clean and accountable.

The 20-file limit is not a concern. Julian says his team never needs more than 20 files attached to a custom GPT.

08Tips, Limits, and Mistakes to Avoid

Free models often get rate limited. Omniroot fixes this by switching between different free APIs automatically, so your agents keep working.

Qwen 3.6 is no longer free, so do not rely on it. Use one of the five other free models or plug in a subscription you already have.

For Obsidian, make sure your agents export all memories, not just part of them. The system works best when everything is organized into folders and linked properly.

With SEO automation, start with the simple skill before moving to the full agent OS setup. The complex version takes longer to configure but is more powerful.

For custom GPTs, avoid hosting your own model. Sharing a company account and links is faster and simpler for teams of any size.

09Who Agent OS Is For and Why It Matters

Agent OS is for anyone building with AI automation, from beginners to experts. It helps you link agents, manage memory, and scale your work.

Julian emphasizes that learning AI is a long journey, and we are all early. Starting now gives you a head start and lets you connect with others who are also learning.

He recommends joining the AI Profit Boardroom community to ask questions, get tutorials, and join weekly coaching calls. The community shares wins, testimonials, and practical help for growing with AI.

Using agent OS with Obsidian becomes your competitive advantage. Every day the memory improves, the links get better, and your AI agents become more useful to you.