Tech notes

Not much to mention on the tech side these days since I’m mostly doing acquisition. Nevertheless. I bought and installed a homeserver, I rescued files from an encrypted SSD and I created a DESIGN.md Homeserver I finally bought and installed a homeserver. I bought a refurbished HP EliteDesk 800 G6 Mini and it’s a dream. 16 GB of RAM, 512GB SSD harddisk and 12 Intel Core i5s for ~400 Euro. Low power usage, silent and very small form factor. ...

I built and deployed the MVP version of Walter, a quarterly Environmental Supply Chain Risk Radar. It was fun and I learned a lot about skills, MCP, agents, token usage, context, sessions, harnesses, Cloudflare Workers & Pages and AI workflow management. Product Development & GTM; Skills & MCP I never coded or designed anything using harness so I did some research on what people are using. The Matt Pocock skills looked promising and they certainly lived up to the hype. I used wayfinder to explore ideas, narrow scope and focus on a first outcome. Then I used the grill-me skills to get rid of ambiguous terms and make clear decisions. The outcome of the skills are documented in my Obsidian vault in a project (I’m using the PARA method). The tasks I manage with Linear. ...

Let’s get ready to build! I installed OpenCode and Pi and was amazed by how smooth the installation goes. Really quick and nice. And after being away from the command line for 10 years, it’s so nice to see the modern TUI animations. Wonderful! I want to use the Claude model family for production, so I signed up for a Pro account. Works great. Installed the Claude Code cli client, authenticated and we’re all good. Then I go to OpenCode to switch the model to Claude Sonnet. It asks me for an API key. No panic, just go to the web dashboard and look around in the settings. Nothing. Checked the documentation and found Anthropic has 2 kinds of subscriptions: chat / cowork (including code) and developer (platform). And the accounts don’t seem to be linked. AND THAT PISSES ME OFF. I’m okay with paying for stuff. But I am not okay with this intransparant bullshit. I haven’t figured out what to do yet. I don’t feel like also paying for an API key. ...

Some leftover chores. I setup Calendly as my meeting scheduler, with Google Meet for web conferencing (I contemplated Jitsi but that does not integrate in Calendly as smoothly); I installed the Excalidraw plugin for Obsidian. It’s insane. Especially in combination with a tablet and a stylus pen. Madness; Hugo supports emojis short codes: 🦊 Hugo emojis quick reference 🦉 ; Prompt engineering to skill engineering to context engineering to loop engineering to graph engineering. And I probably missed some. There’s a new MUST HAVE skill every month in the wonderful world of LLMs. Apparently LangChain and LangGraph are the tools to use. So I wachted the LangChain vs LangGraph video from the IBM series on AI. I like their AI videos. (spoiler: LangChain and LangGraph are both graphs. But LangChain is a directed and acyclic graph, a DAG). Update on Calendly: ...

Bootstrapping This Hollow Earth. I want to know if the AI hype is real. To find out, I decided to start using AI in my day to day workflow. The idea is as follows: I am consulting about technology, product development, organisations and environmental supply chain risk. Great, but there are many people doing that and claiming to be good. How do you know if I’m for real or full of shit? By show, don’t tell. I build products and show how I do it. To build (design, engineer, ship, market, sell) the products, I will use AI. Using AI to build products allows me to learn and evaluate. I write about my adventures and publish them to exploit the teachers paradox. So, to get us started, we need 2 things: ...