Matchids Content

Content management for Matchids: book metadata, artwork metadata, categories, age groups, languages, and authors — the structured data that describes what's in the library, kept separate from the code that serves it.

What's here

books/
  <book-id>/
    metadata.json     Validated against schemas/book.schema.json
schemas/
  book.schema.json
  artwork.schema.json
  categories.json      Real category list — starts empty on purpose
  languages.json
artwork/
  <artwork-id>/
    metadata.json      Validated against schemas/artwork.schema.json
docs/
  STORAGE.md            Why PDFs/images don't live in this repo, and the intended pipeline

Every folder currently contains one _sample- entry — a worked example of the expected shape, clearly not real content. See each sample's README.md for how to add a real entry the same way.

Why this is separate from matchids-database

The database stores whatever's live in production. This repo is where the content team works — draft, in-review, and published entries all live here with a status field, and only published ones with a real storage reference are meant to flow into the database (see docs/STORAGE.md for the intended pipeline).

No large binaries in Git

Book PDFs and images are never committed here — see docs/STORAGE.md. Cover images and Kids Art assets live in matchids-images; PDFs go to object storage, referenced by key.

Validating content

npm install
npm run validate

Runs automatically in CI on every PR (.github/workflows/validate.yml) — a malformed metadata.json fails the build rather than reaching production.

No invented content

Every real entry must be something Matchids actually has permission to publish — no invented authors, no placeholder statistics, no content copied without rights. The _sample- entries are clearly marked and should never be treated as real.