This Website

What it is#

You're looking at it. A personal blog + portfolio that renders to plain HTML at build time — no server, no database, no runtime bill — and still ships the full modern checklist: per-page metadata, build-time Open Graph images, structured data, RSS, sitemap, dark mode, and an llms.txt for AI crawlers.

Why I built it#

I wanted a home for writing and projects that I could not outgrow: something fast everywhere in the world, free to run indefinitely, and readable by both humans and the AI assistants people increasingly ask first.

How it works#

• Content lives as Markdown files with frontmatter in content/. Adding a post is a file commit; the list pages, tag pages, RSS, sitemap and llms.txt all regenerate automatically.
• Build runs next build with output: 'export', producing an out/ folder of static files.
• Markdown pipeline uses remark + rehype with Shiki for dual-theme (light/dark) syntax highlighting, heading anchors, and an auto table of contents.
• SEO/GEO layer emits JSON-LD (WebSite, Person, BlogPosting, BreadcrumbList, optional FAQPage), canonical URLs, and OG images rendered at build time with next/og.

Notes & trade-offs#

Static export means no server-side search or analytics — which is a feature as much as a limitation. Search is delegated to engines, comments can be added via giscus (GitHub Discussions), and privacy-friendly Cloudflare Web Analytics slots in with one config value.

Read the full write-up in How This Site Is Built.