API reference

SEO metadata

article.seo: everything the head of an article page needs, precomputed.

Field	Type	Goes to
`metaTitle`	`string`	`<title>`. About 60 characters at most.
`metaDescription`	`string`	`<meta name="description">`. About 160 characters at most.
`canonicalPath`	`string`	`<link rel="canonical">`, resolved against your origin. Site-relative, under your base path (Articles live under), no trailing slash: `/blog/my-article`.
`keywords`	`string[]`	The queries the article targets, for your own reporting.
`openGraph`	`{ title, description, image?, type? }`	`og:` tags. `image` is the cover.
`author`	`{ name, role?, url?, sameAs? }?`	A byline. Present only when you name an author (Settings → Site and URLs); otherwise the byline is your brand.
`jsonLd`	`object`	One `<script type="application/ld+json">`, serialized as is.
`readingTimeMinutes`	`number?`	Derived from the blocks.

Structured data

jsonLd is one schema.org Article, built from the published blocks. It can also contain:

an FAQPage under hasPart, when the article has an FAQ;
a HowTo under mainEntity, when the article teaches a procedure in numbered steps;
an ItemList under mainEntity, for any other numbered sequence of sections.

The example article shows the FAQ case. Don't add your own Article markup next to it.

Serialize it with < escaped, so no string in the data can close the script tag. The JsonLd component does exactly that:

JSON.stringify(article.seo.jsonLd).replace(/</g, "\\u003c");

canonicalPath comes from your base path, set in Dashboard → your site → Settings → Site and URLs → Articles live under, and so does every link between articles. Set it before your first articles. A change applies to articles written or rewritten afterwards; existing articles keep their old path. Only related follows at once, because it's computed on each read.