nuxt-content

Use queryCollection with the collection name, then chain where, order, select, limit, and all or first. Wrap initial render queries in useAsyncData so Nuxt transfers the result in its payload, and pass the event as first argument in Nitro server handlers.

Use defineContentConfig with defineCollection, choosing type page for route-mapped files or data for structured records. Attach a Zod or Valibot schema, source globs, and optional indexes; run npx nuxi typecheck to verify generated types.

Yes, collections can pull from remote Git repositories with authentication or from APIs via defineCollectionSource with getKeys and getItem. Keys must stay stable across builds, and credentials belong in environment variables.

Replace queryContent with collection-scoped queryCollection and use the queryCollection equivalents for navigation, surroundings, and search. Add an explicit catch-all page, render through ContentRenderer, and use .navigation.yml with the v3 built-in fields.

Empty queries usually mean the collection source glob does not include the file, such as .navigation.yml being excluded by a Markdown-only pattern. Confirm the source include and exclude patterns, then run npx nuxi cleanup and restart if generated content is stale.

Nuxt Content supports SQLite (default), Cloudflare D1, PostgreSQL, libSQL, and PGlite adapters configured under content.database. Static builds package content into the app, while server targets restore the generated dump on first query.