Astro Content Collections explained

This site doesn’t store posts in a database. They’re Markdown files with a Zod schema — Astro Content Collections handle loading, validation, and types.

Two collections

Collection	Location	Routes to
blog	src/content/blog/	/blog/<slug>/
pages	src/content/pages/	/<slug>/

Blog posts require title, description, pubDate, and optional tags and draft. Standalone pages only need title and an optional description.

The schema lives in one file

src/content.config.ts defines both collections. When you add a field — say author on blog posts — update the schema, then add the field to frontmatter. astro check catches mismatches immediately.

Drafts stay private

Set draft: true while you’re iterating. Production builds filter drafts out of listings and routes. Flip to false when you’re ready to publish.

Ask Claude to extend it

Add an optional author field to blog posts and show it on the post page.

Claude can update the schema, frontmatter on existing posts, and the blog/[slug].astro template in one pass. That’s the payoff of collections + AI editing: structure stays strict, changes stay fast.