Welcome to the popcut blog
A quick tour of how this blog is built: MDX articles, automatic SEO, and an AI-assisted writing pipeline.
This is the first post on the popcut blog. It exists to confirm the plumbing works — and to show what every article here gets for free.
Written in MDX
Every post is an .mdx file in content/blog/. That means posts live in version control next to the code, review like code, and can embed React components when a plain paragraph won't do.
Frontmatter at the top of each file drives the metadata:
- title and description become the page title and meta description
- keywords feed the SEO tags
- tags group related posts
- draft keeps work-in-progress out of listings and the sitemap
SEO without the busywork
You don't have to think about any of this per post — it's handled centrally:
- Canonical URLs and Open Graph / Twitter cards
BlogPostingJSON-LD structured data for rich results- Automatic inclusion in
sitemap.xmland a sensiblerobots.txt
An AI-assisted content pipeline
New drafts can be generated from a single topic prompt:
pnpm generate:article "How to pick a short-form video editor" \
--keywords "video editor,short form,clips"
The generator drafts a full article with frontmatter, saves it as a draft, and leaves the final judgment to a human. Review it, tighten the voice, flip draft to false, and ship.
That's the whole loop: prompt, review, publish.