Skip to content
AEO Toolkit AEO Toolkit

Articles with a named, linkable author

Agent skill

aeo-article-author-schema · source on GitHub · install with /plugin install aeo-toolkit --marketplace Advance-Labs/aeo-toolkit (all install options)

An agent loads it when: Mark up articles and blog posts with Article, BlogPosting or TechArticle JSON-LD whose author is a real Person linked to an author page, with datePublished, dateModified and a publisher that points at the organization. Use when someone asks to add article schema, author markup, byline or author bio schema, fix “article and author schema missing”, “author not identified” or “Person schema” audit findings, or improve E-E-A-T attribution for blog and docs content. Covers aeo.article-author-schema, eeat.expertise.person-schema and eeat.expertise.article-markup. Uses the aeo-visibility MCP server if connected; works without it.

Audit rules it owns: aeo.article-author-schema, eeat.expertise.article-markup, eeat.expertise.person-schema

Article markup says “this page is a piece of writing”; the author node says who wrote it and where that person can be verified. Together they give an engine something to attribute a quote to. The markup must describe what is visibly on the page: a real byline, real dates.

When this matters / when it doesn’t

It matters for blog posts, guides, docs articles, news and research: pages where a named person or the organization stands behind the content.

It does not apply to homepages, product pages, pricing, landing pages or tool pages. Marking those up as Article to satisfy an audit is misleading markup. The audit knows this: aeo.article-author-schema is skipped when a single-page audit targets the site root. If the site has no editorial content at all, the honest answer is “this does not apply”, not “add a fake blog post”.

An author is not required for the markup to be valid. Google’s Article documentation lists no required properties. author is recommended, and it can be an Organization when the content is genuinely published by the company with no individual byline. Do not invent a person to fill the field, and do not credit a real person who did not write the piece.

AI-generated or ghostwritten content needs a human decision about who is credited. Ask the user; do not choose a name.

Detect

RulePasses when
aeo.article-author-schemaAcross the crawled pages, at least one Article-family node exists and at least one Person node exists.
eeat.expertise.article-markupAt least one Article-family node exists.
eeat.expertise.person-schemaAt least one Person node exists.

What that really means:

  • “Article-family” is Article, NewsArticle, BlogPosting, TechArticle, ScholarlyArticle, Report, APIReference and the other schema.org subtypes of Article. TechArticle counts for presence.
  • Detection is at any nesting depth, so author: { "@type": "Person" } inside the Article satisfies the Person check.
  • The rules are site-wide ORs, not per-page. A Person anywhere (a founder on the About page) plus an Article anywhere passes aeo.article-author-schema, even if no article is attributed to anyone. The audit passing does not prove the author is linked to the article. Check that yourself.
  • The toolkit’s validator requires headline on Article, NewsArticle and BlogPosting, and name on Person. A missing one marks the item invalid (tech.structured-data-valid). TechArticle has no required-property check in the validator, so a headline-less TechArticle is not flagged. Include headline anyway.

With the MCP server: analyze_website_aeo on a sample article URL (not only the homepage), then read the three rules and the structured-data detail.

Without it: view-source an article, find the ld+json blocks, and answer these in order: is there an Article-family node; does it have an author; is the author a Person with name and a url or sameAs; does that URL resolve to a real author page; are both dates present and ISO 8601; does publisher reference the Organization @id.

Decide

SituationDo this
Posts have a visible byline linking to an author pagePerson node with url to that page and sameAs to external profiles.
Byline exists, no author pageCreate a short author page (name, role, bio, links) before marking it up; or link the best existing profile via sameAs. Ask the user which.
Posts are published as the company, no bylineauthor is the Organization ({"@id": ".../#organization"}).
Multiple authorsOne author array entry per person. Never one entry with “A, B”.
Guest postThe guest is the author; the site is the publisher.
Docs or API reference pageTechArticle (or Article). Same author rules.
Page is a changelog or release noteArticle is fine; do not use NewsArticle unless it is news.
Dates are unknownOmit them rather than guessing. Do not set dateModified to today on every build.
Homepage or landing pageDo nothing. Not an article.

Implement

Shape, independent of stack:

  1. One Person node per author, defined once on that author’s page with a stable @id (https://example.com/authors/jane-doe#person). Give it name, url (the author page), sameAs (LinkedIn, GitHub, personal site, ORCID if relevant), and jobTitle and worksFor if true and shown on the page.
  2. Each article references that person by @id in author, and the organization by @id in publisher. Also set mainEntityOfPage to the page and isPartOf to the WebSite.
  3. datePublished and dateModified in ISO 8601 with a timezone offset (2026-11-13T09:00:00-05:00). Google recommends the offset; without it the timezone of Googlebot is assumed. The dates must match the visible ones.
  4. image: relevant article images, crawlable URLs, ideally in 16:9, 4:3 and 1:1.
  5. Generate all of it from the same post data that renders the byline, so they cannot diverge.
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "BlogPosting",
      "@id": "https://example.com/blog/answer-engines#article",
      "headline": "How answer engines choose what to cite",
      "description": "A one-sentence summary that matches the meta description.",
      "mainEntityOfPage": { "@id": "https://example.com/blog/answer-engines" },
      "url": "https://example.com/blog/answer-engines",
      "datePublished": "2026-11-13T09:00:00-05:00",
      "dateModified": "2026-11-20T14:30:00-05:00",
      "image": ["https://example.com/img/answer-engines-16x9.jpg"],
      "author": { "@id": "https://example.com/authors/jane-doe#person" },
      "publisher": { "@id": "https://example.com/#organization" },
      "isPartOf": { "@id": "https://example.com/#website" }
    },
    {
      "@type": "Person",
      "@id": "https://example.com/authors/jane-doe#person",
      "name": "Jane Doe",
      "url": "https://example.com/authors/jane-doe",
      "jobTitle": "Head of Research",
      "worksFor": { "@id": "https://example.com/#organization" },
      "sameAs": [
        "https://www.linkedin.com/in/janedoe",
        "https://github.com/janedoe"
      ]
    }
  ]
}

The Organization node (/#organization) and WebSite node are defined on the homepage; see aeo-organization-entity. If a referenced @id is not defined on the same page, a validator cannot see what it points to, and some engines will not follow it. Where that matters, include the Organization node in the same graph on the article page too, with identical values.

Google’s author guidance, which the markup must follow:

  • author.name is only the name. No job title, no “Posted by”, no honorifics, no publisher name; those have their own properties (jobTitle, honorificPrefix, publisher).
  • Use Person for people and Organization for organizations. Do not use the wrong one or Thing.
  • List every author shown on the page, each as its own entry.
  • author.url should link to a page that uniquely identifies the author (author page, social profile). sameAs is accepted as an alternative and Google understands both when disambiguating. For an author page on your own site, Google suggests marking it up with ProfilePage structured data; that is optional and has its own documentation.

Author page: a real, crawlable page at the url, with the person’s name, a short bio, a photo and links to the posts. The Person node is defined there. It should be linked from the bylines, in the sitemap and not noindex.

Stack recipes: references/nextjs.md, references/astro.md, references/wordpress.md, references/static-html.md.

Verify

  1. Re-run analyze_website_aeo on an article URL (or the validator on its HTML).
  2. The three rules above should be pass. If only the Person rule fails, the author is a bare string ("author": "Jane Doe"); convert it to a Person.
  3. Rich Results Test and Schema Markup Validator on the article URL: no errors; warnings for missing recommended properties are acceptable.
  4. Open the page and compare, field by field, to the markup: headline, byline, dates, image. Any mismatch is a defect even if the validators are happy.
  5. Click the author url; it must return 200, be indexable and show the same name.

Pitfalls

  • "author": "Jane Doe" as a string. Valid for some consumers, but it is not a Person node, so the Person rules fail and there is nothing to link.
  • Author pages that 404, are noindex or are blocked in robots.txt.
  • dateModified bumped on every deploy. Engines and users read it as a freshness claim. Change it only when the content changed. See aeo-content-freshness.
  • Dates without a timezone offset shift by hours in other zones.
  • Two Article blocks on one page from a theme and an SEO plugin, with different authors or dates. See aeo-structured-data-validation.
  • A whole site marked NewsArticle. Use it for news. Article or BlogPosting otherwise.
  • Fake authors, stock photos, invented credentials. This is the opposite of E-E-A-T and violates Google’s guideline against misrepresenting who is behind the content.
  • Marking up the homepage as an Article to pass the rule.

Sources

AEO Toolkit is an open-source instrument by Advance Labs.