lans-h.cc site official launch

• by Lans Hung

I don’t usually share my findings or my development process publicly. My assumption was that there are people out there far smarter than me, and whatever I had to say, they had already said it better.

The skill set that defines a modern engineer has moved with the introduction of LLMs. I think people like us may have gained the most from that shift — the ones who try to think deep and broad at the same time, but hold very little in short-term memory. An LLM holds the complexity better than human does.

Around 2017, two or three years into Oracle, I started a habit of keeping daily work log and a running set of notes (A lot of them were commands I kept having to look up. LLM writes those much better than I do now 🙂). Computer engineering spent decades teaching people to speak the language of machines. Much of that layer now sits behind natural language. Memorizing syntax used to take up a real share of what a software engineer does, and far less of the work goes there now. The detail work at the end now took the longest, but it turned out to be what did the most for how I grew as an engineer. That whole workflow is now AI-assisted end to end. What changed the most is where my time goes, and how little it costs me to share something once I’ve found it.

notes.md, 2017 — the text search section. The internal identifiers are masked.

notes.md, 2017 — teaching myself here strings and here documents.

That is the reason the site exists.

  • platform is the piece I care about most: one shared Kubernetes cluster carrying several independent workloads, with a namespace per app, a shared Postgres with a least-privilege role per app, a wildcard certificate that keeps TLS a platform concern, not an app one, and one signed deploy path every repo hooks into. Onboarding a new app comes down to four additive steps, so the next one costs about what the last did. The topology diagram is on the site.

    The platform project card

  • Alongside it are three Claude Code plugins I’ve published — upward-mobility, upward-stats, and explain-to-lans — and a number of projects still in progress.

    The upward-mobility project card The upward-stats project card The explain-to-lans project card

  • I’ve also been sitting on a set of findings about Claude Code models and autonomous agentic workflows, and I hope to write those up before long (coming soon).

More to come as I find it.

Lans Hung