SEO & JSON-LD

Every field you need for on-page SEO and structured data ships inside the Article. There is nothing to compute and nothing to fetch separately — map the fields to meta tags and JSON-LD.

The meta fields

Field	Use it for	Notes
`seoTitle`	`<title>`	Always present — falls back to `title` server-side. Use it directly.
`seoDescription`	`<meta name="description">`	`string \| null`. When `null`, omit the tag and let the search engine derive one.
`canonicalUrl`	`<link rel="canonical">`	`string \| null`. `null` only when the site has no URL set.
`inLanguage`	`<html lang>` / JSON-LD `inLanguage`	BCP-47 tag, e.g. `"en"`.
`formatType`	JSON-LD `@type` hint	schema.org type, e.g. `"Article"`.
`tags`	`keywords` / topical metadata	May be empty.

Because seoTitle is guaranteed and the two nullable fields are explicitly typed, you can emit clean head tags without guesswork:

const head = {
  title: article.seoTitle,
  description: article.seoDescription ?? undefined,   // omit when null
  canonical: article.canonicalUrl ?? undefined,
  lang: article.inLanguage,
};

The hero image

When present, image is everything you need for an Open Graph card and a layout-shift-free render in one object.

interface ArticleImage {
  url: string;            // og:image and your <img src>
  alt: string;            // <img alt>
  width: number | null;   // set <img width/height> from these
  height: number | null;
  srcset: string | null;  // <img srcset> for responsive sources
}

Use image.url for og:image (and twitter:image).
Set <img width> / <img height> from image.width / image.height so the browser reserves space and you avoid cumulative layout shift.
Drop image.srcset straight into <img srcset> when it’s set.

image is null when the article has no generated cover — fall back to a site-level default OG image, or omit the card.

author.kind selects the JSON-LD shape

The author’s kind is the switch between a Person and an Organization in your structured data.

`author.kind`	JSON-LD `author` `@type`	When
`"real"`	`Person`	A real human byline. `sameAs` carries their profile URLs.
`"persona"`	`Organization`	A brand byline rather than an individual.

function authorLd(author: Article["author"]) {
  if (author.kind === "real") {
    return {
      "@type": "Person",
      name: author.name,
      jobTitle: author.jobTitle ?? undefined,
      image: author.avatarUrl ?? undefined,
      description: author.bio ?? undefined,
      sameAs: author.sameAs.length ? author.sameAs : undefined,
    };
  }
  return {
    "@type": "Organization",
    name: author.name,
    sameAs: author.sameAs.length ? author.sameAs : undefined,
  };
}

A complete Article JSON-LD node then assembles from the same fields:

const articleLd = {
  "@context": "https://schema.org",
  "@type": article.formatType, // e.g. "Article"
  headline: article.title,
  inLanguage: article.inLanguage,
  datePublished: article.publishedAt,
  dateModified: article.updatedAt,
  description: article.seoDescription ?? undefined,
  image: article.image?.url,
  mainEntityOfPage: article.canonicalUrl ?? undefined,
  author: authorLd(article.author),
};

What to reach for next

The article shape The full field-by-field reference behind these tags.

The content feed Where the SEO fields come from.

Instant updates Keep canonical and metadata fresh as articles change.