⚙️ How this blog works

This site is a deliberately simple markdown blog. There is no CMS, no database, and no fancy framework — just files in a GitHub repository and a GitHub Action that turns them into pages.

The idea

Write a post as a .md file, push it, and the site updates. That is the whole workflow.

Under the hood it uses my fork of github-pages-blog-action (v0.1.0): on every push to main/master, the Action builds HTML from markdown and deploys it to the gh-pages branch.

What lives in the repo

• posts/ — articles; future date or draft: true stay unpublished
• drafts/ — unfinished posts that never go live
• about.md — the about page
• site.json — title, SEO, social links, site URL, favicon, OG image
• static/ — files copied to the site root (theme overrides, images, favicon, og.png)

Each post starts with frontmatter:

---
title: "How this blog works"
date: 2026-08-11
description: "Optional excerpt for SEO/social"
---

The date is both the display/sort date and the publish gate: if it’s in the future (UTC), the Action skips that file. A daily cron redeploys around midnight UTC so scheduled posts go live without a manual push. Use draft: true or the drafts/ folder for unfinished work.

Local preview

npm install
npm run preview

Use npm run preview:drafts, npm run preview:scheduled, or npm run preview:all when you need those.

Theme

The Action ships an amber dark theme (date under title, zoom controls, Share). This repo can still override CSS/JS via static/css/theme.css and static/js/site.js because static files are copied last.

Why keep it this simple

Markdown in, website out — nothing more. When the urge to over-engineer shows up, the answer is still: add another .md file and push.