An Agentic Operating System (Agent OS) is a unified platform that consolidates multiple AI agents into a single interface, enabling automation of tasks like outreach, content creation, and video production without switching between tools. The system operates on five layers: foundation (agents in one place), memory (automated context storage), tools, models, and actions. The memory layer is the critical differentiator, as it automatically updates and connects all agents to user context, making the system feel intelligent and responsive. Fable 5, developed by Anthropic, enables non-technical users to build such systems through natural language descriptions, with access available on paid Claude plans.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
Fable 5 Agentic OS Makes Automation SCARY
Added:Fable 5, Anthropic OS, makes automation scary.
What if one system could run almost everything you do all day? No tabs, no logging in, no jumping between 10 different tools, just one screen where every AI agent works for you. Sounds fake, right? It's already real, and it's a little scary how well it works. Hey, I'm the digital avatar of Julian Goldie, and I help people actually learn and use AI tools in their real work. Today, I want to walk you through my Agent OS.
It's an agent operating system I built myself, and the big new piece is that I'm now building it with Claude Fable 5 right inside my Agent OS. Stick with me.
By the end, you'll see how I run outreach, content, and videos from one place, plus the five layers that hold it all together. So, let me start with the big picture. My Agent OS is mind-blowing, and I don't say that lightly. It puts all of my AI agents in one place, so I can automate almost anything I do without hopping around. I coded a big part of the new version with Fable 5 the moment it dropped, and I'll show you what I built as we go. Here's the simple truth. You can use one AI in a terminal and another in a browser tab, but you'll never keep up with a full system like this where every agent is one click away. Here's a real example. I use this to build a full outreach system that brings the right people toward the AI Profit Boardroom. I describe the kind of people I want to reach, and it builds a list, drafts the emails, and runs the whole outreach flow for me. I can check what got sent and what came back all in one spot. That's the kind of quiet, steady work that keeps the AI Profit Boardroom growing while I focus elsewhere. Now, let's talk about trending news because this part is wild.
I can pull the latest trending AI news, and it refreshes on its own every 24 hours. From there, my agents draft social posts and a full blog post, and I publish it to my site in basically one click. So, here's how I use that for the AI Profit Boardroom. I take a hot trending topic, let it write a clean blog post around that keyword, and push it live to pull search traffic toward the AI Profit Boardroom. That means fresh content and new eyes on the AI Profit Boardroom without writing every word by hand. Next up is my voice in the things. This is Hermes Jarvis. It's a voice-activated version of my Hermes agent, so I can just talk and build in real time. I can see the full transcript of what I said, open it up in what I call war mode, and watch it start building. This is what I use to talk out a new idea for the AI Profit Boardroom out loud, like a page or a small tool, and watch it come to life while I'm still speaking. It feels less like typing and more like thinking out loud with a fast helper. Okay, now let me slow down and show you why this all holds together. My Agent OS works in five layers. The first layer is the foundation. That's all of my agents living in one place, so nothing is scattered. The second layer is the one people get most excited about. It's my memory. I call it a memory galaxy, and it's fully automated by my agents. Every useful thing I've ever worked on across my agents lives in there. My vision, my goals, my plans, my projects, the stuff I'm actually working on day-to-day.
Every piece is stored as a little node in that memory. And here's the part I love. It updates itself. As I work, new things get added to the memory on their own, so I don't have to stop and save everything by hand. Then that memory is plugged straight into every agent I have. So, when I ask any agent to do something, it already knows my context.
It knows what I'm building and where I'm heading. The other three layers sit on top of that, connecting my tools, my models, and the actions I want to run, so the whole thing moves as one. Once you see it working, it's hard to go back to bouncing between random apps that forget you the second you close them.
This memory layer is honestly the difference between a pile of clever tools and a real system that feels like it knows you. It's the quiet engine under everything else, and it's a big reason the rest of the Agent OS feels so smooth. So, if you want my exact Agent OS setup, here's where to get it. It's inside the AI Profit Boardroom. You get the full Agent OS zip file, ready to install, so you're not starting from a blank screen. On top of that, I built a full 30-day road map that walks you through it step by step, so you know exactly what to set up first and what to do next. Inside the AI Profit Boardroom, you also get live coaching calls where you can ask me about your own setup, plus walk-throughs and the prompts built around this exact system. It's the fastest way to go from watching me do this to running your own version of it.
Now, the fun creative agents. I've got a movie creator that builds a video in one click. I've got a thumbnail studio that makes thumbnails that honestly look better than most people can design by hand. I've got a music agent for quick background tracks, and I can even make short clips for free using a notebook style tool built in. Here's how I put these to work. I use the thumbnail studio and the clip agent to make eye-catching thumbnails and quick videos that pull the right people toward the AI Profit Boardroom. Good thumbnails and clips get people to stop and click, and that click is the first step into the AI Profit Boardroom. Then there's my video agent. I type what I want, I say how long it should be, up to about 10 minutes, and I pick a presenter or just a voiceover. I choose which agent handles the B-roll, then it writes the script, does the research, and builds the whole thing. I use this to spin up a full video introducing the AI Profit Boardroom with the script, the research, and the B-roll all handled in one flow.
So, instead of a whole day editing, I get a finished video that sends people toward the AI Profit Boardroom. I also automate my whole website and search site. I take my own search data, plug it into a keyword, and start publishing content across my site in one click. I honestly haven't logged into a website by hand in ages because my agents handle it. This is how I keep fresh, search-friendly pages going out that point people to the AI Profit Boardroom.
If you're still posting every blog by hand, this part gives you your time back. Now, you might be wondering, how do you even start building something like this? It's simpler than you think.
You open Claude, select Fable 5, and describe what you want in plain words.
That's it. Describe the piece you want like a mission control screen with your agent and let it build. And a quick heads-up on Fable 5 access because I want to keep this accurate. Right now, Fable 5 is included on the paid Claude plans up to about half of your weekly usage, and that window runs through July 19th. After that, it moves to usage credits. So, if you want to build with it while it's included, that's your window. And here's something I really want you to hear. I'm not a coder. I don't speak HTML. And a huge number of the people building this with me aren't coders, either. That's the whole point.
People who've never written a line of code are setting up their own Agent OS and getting it running. If that sounds like you, you're exactly who this was made for. The other thing I love is how fast I can plug in new tools. Anytime something genuinely useful drops, I add it. When Gemini 4 got a big speed boost with MLX, local models suddenly became actually useful, and I plugged it in the same day. When I spot something handy on GitHub, like a movie or design tool, I just wire it in. So, my system keeps getting better without me rebuilding it from scratch. Now, a fair question.
Doesn't all this use a lot of tokens? It can, but there are smart ways around it.
I lean on free and local models where I can. I use models I already have through a CLI, plus free APIs and open models inside the system. So, a lot of this runs for free or close to it once you set it up right. That's a big reason it stays practical, not just a cool demo.
You might also ask if I actually tested any of this. I did. I built my own setup I call Goldiebench, where I put a lot of builds through many different models and tasks, then sort them by what each one is best at. Fable 5 stood out for building things, especially anything with code or visual polish. It's quick and clean. The big win for me is the speed of going back and forth. When I'm tweaking a build, I don't want to sit and wait. That leads me to the other models I keep in my Agent OS. I've got a mixture of agent setups in there, too, like Fusion and my Hermes mixer. They can build some seriously impressive stuff, but they can be slow, so you might wait several minutes for one reply. Fable 5 lets me go back and forth in seconds, which is why I reach for it when I'm building fast, so I can pick the right tool for the moment. So, why does all of this matter? Because the real magic isn't any single agent. It's that everything lives in one place, remembers my context, and sits one click away. That's what saves me hours every single day. If you're going to try this yourself, here are my honest tips. Start small. Build one agent you'll actually use, then add to it a little each day.
Lean on that memory layer early, so your system learns your context from the start, and only plug in tools that are genuinely useful, not every shiny thing you see. Keep it simple, and it'll keep giving you time back. If you want the full process, SOPs, and 100 plus AI use cases like this one, join the AI Success Lab. Links 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 ready to actually build your own Agent OS, come join me inside the AI Profit Boardroom. You're about to run into the same questions I did when I started.
Which tool do I plug in first? How do I set up my memory layer? How do I keep my tokens low? That's exactly what we solve together. You get live coaching calls to ask about your own setup, step-by-step tutorials, the full 30-day roadmap, and the prompts I actually use, all with over 4,000 members building right alongside you. Come get it at aiprofitboardroom.com.
I'll see you on the next one.
Related Videos

TOP 15 Data compression Interview Questions and Answers 2019 Part-2 | Data compression | Wisdom jobs
wisdomjobs
281 views•2019-06-28

CTS 158: 802.11w Management Frame Protection
ClearToSend
4K views•2019-02-04

NDSS 2019 Send Hardest Problems My Way: Probabilistic Path Prioritization for Hybrid Fuzzing
NDSSSymposium
496 views•2019-04-02

How realistic is Cities: Skylines?
CityBeautiful
159K views•2019-02-14

GUIs & TUIs: Choosing a User Interface for Your Python Project | Real Python Podcast
realpython
2K views•2025-04-04

The OSI Model - Explained by Example
hnasr
225K views•2019-05-12

Cloud Computing - Introduction
elithecomputerguy
98K views•2019-10-07

From Traveler's Dilemma to Dynamic Routing | Demystifying Networking
IITBombayJuly
5K views•2019-08-04
Trending

WOW! Judge TURNS THE TABLES on Trump in His OWN $10B LAWSUIT!!!
MeidasTouch
197K views•2026-07-23

Playstation NO DISC/NO BUY Fight Is Over...
DavidJaffeGames
4K views•2026-07-23

Steam and Xbox Just Dropped The Hammer On PlayStation
OhNoItsAlexx
9K views•2026-07-23

Americans Confused in Australia for 17 Minutes Straight
IWrocker
17K views•2026-07-23