GitHub profiles have become something of a personal homepage for developers. Mine at github.com/abuxton has been through a fair few iterations, and I thought it was worth walking through what’s there, how it got there, and how it connects to the rest of how I work.

The Profile Repository

GitHub’s special profile repository — the one where your README.md appears on your profile page — is abuxton/abuxton. It is labelled plainly: “profile repository holding page”. Understated, but accurate.

The README is my attempt at a genuine, no-nonsense introduction. It opens with a multilingual welcome — a small reminder that the internet is global — and moves straight into who I am and what I do: a professional consultant focused on helping people solve their own problems, not solving problems for them. Teaching people to adopt new technologies — AI, agentic workflows, IaC, DevOps — is the throughline.

There is a line in there I like: “My life is a constant round of ‘could you, should you, and would you?’ It’s like ‘kiss, marry, or avoid’ but for technology-related products.” That about sums it up.

The User Manual

One of the sections I find most useful on the profile is the User Manual — a table format borrowed from Cassie Robinson’s original idea. It covers:

  • Conditions I like to work in
  • The hours I prefer
  • How I like to receive feedback
  • Things I need
  • Things I struggle with
  • Things I love
  • Other things to know about me

The practical upshot: I’m an early bird, I prefer Slack over screen grabs, and I’m Dyspraxic — which means people points are a real, finite resource. The further into the week it gets, the more introverted I become.

I put this on my profile because working well with people starts with telling them honestly how you work. Most teams spend months figuring this out the hard way. Publishing it upfront saves everyone time.

The profile includes a set of badges: the Awesome badge, Puppet Forge module count, Reddit karma (under my adept2051 handle), GitHub followers, and a pair of license badges for abuxton/.github and abuxton/dbad — the last of which is my port of the Don’t Be A Dick open source licence.

Badges are cosmetic, but they are also a quick signal about where someone is active and what they care about. Puppet Forge presence says IaC. The DBAD licence says I believe open source is better when people are decent about it.

The Reading List

Below the badges is a curated reading list that has accumulated over time. A few highlights worth repeating:

  • People are not resources — a long-held belief that still needs saying.
  • The Scotty Principle — engineers consistently under-scope time and effort; if we applied Scotty’s approach more often, we’d still be wrong, but by less.
  • Do-nothing scripting — the smartest piece I know on the when and whether of automation, not just the how.
  • Tenets of IT — Taoism applied to modern engineering. Underrated.

The Geekcard

The contact section points to the Geek Card — a small npm-based card (npx digitaladept) that outputs contact information in the terminal. This landed via PR #1 — Feat-geekcard in October 2025. It is the kind of small, unnecessary thing that is entirely worth doing.

The Dotfiles Connection

The profile does not exist in isolation. The abuxton/dotfiles repository is what actually configures the environment I work in day to day, and the two are increasingly connected.

Why the Profile and the Dotfiles Belong Together

The profile README is a document about how I work with people. The dotfiles are a document about how I work with machines. The skills framework is, increasingly, a document about how I work with AI agents.

They are all the same conversation, just addressed to different audiences. The user manual in the profile tells humans how to work with me effectively. The skills in ~/.agents/ tell AI agents the same thing — the conventions I prefer, the workflow patterns I follow, the tools I use.

The Ghost Engineer note on the profile captures the underlying point. The work that statistics and commit counts don’t show — the explaining, tutoring, collaborating, communicating — is the most important work. The AI tooling I have been building into my environment is intended to augment that work, not replace it.

Where to Find Everything