How I use AI to build and run my projects.
Inside my setup: OpenClaw, Notion, a Second Brain and the tools that connect the work.
I am a marketer who builds things, including websites, content systems and AI workflows. The useful part is how the tools fit together: remembering decisions, finding the right context, doing the work and checking the result.
I call the operating system around this work Agent OS. My Second Brain holds the knowledge, Notion gives me a place to review it, and tools such as Codex and OpenClaw handle defined tasks. I also built my own agent, Zeus, which I control through Telegram. Here is how I have put those pieces together.
Zeus: my own agent, a Telegram message away
I built Zeus to work across Agent OS from end to end. I can give it direction through Telegram and use it to coordinate the system, carry out work and bring the results back to me.
Zeus is also my day-to-day assistant for life, the universe and everything I need help with. That gives this setup a purpose beyond building websites: I want the knowledge and tools I have put together to be useful in everyday life too.
Zeus is the agent I talk to. Zeus Dispatch is the separate broker that routes work within the setup.
How the setup fits together
- DirectionI talk to Zeus through Telegram to direct work across Agent OS.My own agent
- MemoryObsidian holds notes, decisions and reusable instructions.My Second Brain
- Context and planningqmd finds the relevant notes so the work starts with the right context.The workflow
- ExecutionI use Codex and OpenClaw to turn the plan into working code and content.The tools doing the work
- Checks and reviewI check the result and review completed work in Notion.My command centre
- Output and loopThe finished work goes where it is needed. I save the decisions and lessons in my Second Brain for the next task.Back into the system
The output might be a website, an email or an answer from Zeus. Saving what worked means the next task can build on it.
What that looks like on Sideperk
I use Codex to build and update the website, Fedora for larger builds and tests, and Cloudflare to host it. Listmonk holds the newsletter and mailing list, with Twilio Email delivering each edition.
The research, decisions and useful lessons go back into my Second Brain, ready for the next update.
Where you can see the work
My public project portfolio covers the wider work. At MorpheOS Ltd, I develop IntelligenceOS and software for working with AI agents. These are my own projects, and I have a commercial interest in them. Let's build together covers the services I offer around website and AI workflow work.
My coding tools and model choices
I use ZCode with DeepSeek Pro, GLM and OpenAI, as well as Codex for website work. My guide to ZCode, OpenCode and Codex explains the different interfaces and where model access fits in.
The tools and their jobs
ZCode
Working with several AI model providers
I use ZCode with DeepSeek Pro, GLM and OpenAI. It is one of the coding tools in my setup, alongside Codex.
GLM by Z.ai
Another model option in ZCode
I use GLM through ZCode, alongside DeepSeek Pro and OpenAI.
OpenClaw
Running AI tasks
I use OpenClaw alongside Codex in Agent OS. Codex is my first choice for coding; OpenClaw gives me another way to run tasks and bring the results back into the same workflow.
Notion
Reviewing priorities and completed work
I have built a Notion command centre for priorities, projects, ideas and completed agent work. It presents selected records from the underlying system in a form I can review, with links back to the supporting material.
OpenAI Codex
Building and checking websites
I use Codex to work on website code, investigate faults and turn a brief into changes I can inspect. Sideperk itself is one example: the work includes its pages, content handling and release checks.
GPT Pro
Research and scenario runs on frontier OpenAI models
I have a GPT Pro subscription. I run research and scenarios through the latest frontier OpenAI models, keeping my project work close to the current edge of what the models can do rather than my own assumptions about them.
DeepSeek
Building a tool for repeatable AI batches
I have built my own DeepSeek-based coding and batch tool. I use it for batches of first drafts, outlines and structured content, then review the results together.
Obsidian and my Second Brain
Keeping the decisions behind the work
My Second Brain is the knowledge system around the work. I keep project notes, decisions and operating instructions in Obsidian, and give recurring practices a consistent structure in a Knowledge Catalog. This page began with those records.
qmd
Searching my local knowledge
I use qmd to retrieve the relevant notes from my Second Brain before work starts. It gives my assistants a local search layer across the project records and operating instructions held in Obsidian.
ElevenLabs
Generating voice for audio and avatar work
I use ElevenLabs for voice cloning and generated speech. My recorded Aurora avatar test used an ElevenLabs voice clone.
HeyGen
Making avatar-led video
I used HeyGen Avatar IV in a matched short comparison with the same source image and voice. In that one test it produced a higher-resolution result at a lower recorded credit cost than the Aurora result.
Immich
Managing my photo and video library
I run Immich on my own infrastructure for photo and video management, backed by separate storage and recovery checks.
Colima
Running containers on my Mac
I use Colima as the container runtime on my Mac for local Docker-based development and service checks.
Caddy
Routing my self-hosted services
I use Caddy as the reverse proxy in my self-hosted setup, where it routes named services and manages HTTPS.
Open WebUI
Using local models in a browser
I run Open WebUI on my Fedora workstation as the browser interface for local models and supported model APIs. Its current licence is source-available rather than an OSI open-source licence.
ComfyUI
Building image-generation workflows
I run ComfyUI on demand on my Fedora GPU workstation for node-based image generation workflows.
Audiobookshelf
Listening to audiobooks and podcasts
I run Audiobookshelf in my self-hosted media stack for organising and listening to audiobooks and podcasts.
Calibre-Web Automated
Organising my ebook library
I run Calibre-Web Automated in my self-hosted media stack to organise and process an ebook library.
Cloudflare
Running Sideperk
Sideperk runs on Cloudflare Workers, with D1 storing its editorial records and Access protecting the editor area. It gives this small publication a place to run without a separate web server to maintain.
Listmonk and Twilio Email
Putting The Side Piece in your inbox
I use Listmonk to prepare newsletters and manage lists, with Twilio Email handling delivery. For Sideperk, that has included a reusable email template, a first edition of useful listings and a delivery check in a real inbox.
MXroute
Handling ordinary email
MXroute handles the Sideperk contact mailbox. It keeps reader questions, corrections and replies separate from the newsletter delivery system.
Fedora
The heavy-work GPU machine
I built Project Zeus as a dedicated high-power machine for the heavier side of the work. It runs Fedora Linux with ROCm serving a local Qwen3-8B model, Open WebUI and on-demand ComfyUI, with ZFS storage for the working files. When a larger build, batch of drafts or benchmark-style check needs room to run, the job goes there and the Mac stays free for everyday work. The build is tuned and stress-tested, with recorded baselines in my notes for each component.
- AMD Ryzen 9 7900X, 12 cores / 24 threads, up to 5.6 GHz
- XFX Radeon RX 7900 XT Speedster MERC 310, 20 GB
- 48 GB DDR5-6000 CL28 memory
- MSI MAG B650 Tomahawk WiFi motherboard
- Corsair RM850e 850 W 80 Plus Gold power supply
- NZXT H5 Flow case with a DeepCool AK620 cooler
- Fedora Linux, ROCm and ZFS storage
My background in AI and digital work gives more context.
These are direct links, with no affiliate tracking added to this page. The offer listings cover free plans and eligibility. Contact and corrections are available for changes to a tool or workflow.