Structured data is a small, machine-readable summary of your page that Agentimus prints in the page’s HTML head. Humans never see it, but search engines and AI assistants read it to understand who you are, what a page is about, and how your pages relate to each other — without having to guess from your layout or wording.

Agentimus writes this summary in JSON-LD (JavaScript Object Notation for Linked Data), the format Google, Bing and the major AI answer engines all understand. It follows the schema.org vocabulary, so the labels it uses (WebSite, Person, BlogPosting, and so on) mean the same thing to every tool that reads them.

This page explains exactly what Agentimus emits, when, and how to shape it. You do not need to write any code — everything here comes from your settings.

Why it matters

When a person searches or asks an assistant a question, the tool has to decide what your page is and whether to trust it. Clean structured data gives it a direct, unambiguous answer:

  • You are identified. The assistant knows the site belongs to a specific person or organisation, not an anonymous page.
  • Pages are typed. A blog post is marked as an article; a page is marked as a page; a genuine question-and-answer page is marked as an FAQ.
  • Everything is connected. Each article points back to you as its author and publisher, so an assistant can attribute your work correctly.

That is the difference between an AI answer that says “according to Jane Smith’s guide” and one that quietly paraphrases you with no credit.

Where it appears and when

Agentimus hooks into WordPress at the very top of the page head and prints a single <script type="application/ld+json"> block. Inside that block is one graph — a list of connected nodes — tailored to the page being viewed:

  • On every front-end page, it emits the two site-level identity nodes: your WebSite and your Person/Organization entity.
  • On your front page, it also adds any Service nodes you have declared.
  • On a single post or page (of a content type Agentimus covers — posts and pages by default), it additionally emits the per-page nodes: the article/page node, a breadcrumb trail, and an FAQ node if the content clearly is a Q&A.

By default Agentimus covers posts and pages. If you have opted extra content types in (for example WooCommerce products or a custom post type) under Settings, those single views get the per-page nodes too.

The site identity nodes

These two nodes describe the site as a whole and appear on every front-end request.

The WebSite node names your site and, importantly, declares a SearchAction — a standardised description of your site’s own search box. This is what can let Google show a sitelinks search box (a search field right inside your listing) for your brand.

{
  "@type": "WebSite",
  "@id": "https://example.com/#website",
  "url": "https://example.com/",
  "name": "Example Site",
  "description": "Your site's tagline",
  "inLanguage": "en-US",
  "potentialAction": {
    "@type": "SearchAction",
    "target": {
      "@type": "EntryPoint",
      "urlTemplate": "https://example.com/?s={search_term_string}"
    },
    "query-input": "required name=search_term_string"
  }
}

The name, tagline and language come straight from your WordPress General Settings and Site Language — there is nothing extra to configure.

Your identity: Person or Organization

This is the heart of the graph. It says who stands behind the site. Depending on what you choose on the Identity page, its type is either Person (a human — a blogger, freelancer, author) or an Organization type (Organization, LocalBusiness, Store, and any others an add-on adds).

{
  "@type": "Person",
  "@id": "https://example.com/#identity",
  "name": "Jane Smith",
  "url": "https://example.com/",
  "jobTitle": "Pastry Chef",
  "description": "I write about home baking and pastry technique.",
  "disambiguatingDescription": "Not a commercial bakery or a recipe aggregator.",
  "audience": { "@type": "Audience", "audienceType": "Home bakers" },
  "knowsAbout": ["Sourdough", "Laminated dough", "Chocolate work"],
  "sameAs": [
    "https://www.wikidata.org/wiki/Q42",
    "https://mastodon.social/@janesmith"
  ]
}

Every field here is drawn from the Identity page:

JSON-LD field Identity setting What it does
@type Entity type Person, or an Organization subtype
name Name The person or organisation’s name
jobTitle Role Your role (Person only)
description About A short description of the site
disambiguatingDescription What this is not Tells assistants what not to confuse you with
audience Audience Who the site is for
knowsAbout Areas of expertise The topics you are authoritative on
sameAs Verified profiles Links to your Wikidata entry, social profiles, etc.

The sameAs links are especially valuable: they let an assistant confirm it is talking about you and not someone with a similar name.

Services you offer

If you have listed any services on the Identity page, each becomes a Service node so an assistant can answer “what can I hire this site for?”. Services are opt-in — Agentimus never guesses them from your content — and they ride on the front page.

{
  "@type": "Service",
  "name": "Wedding cake commissions",
  "description": "Custom multi-tier cakes for events.",
  "url": "https://example.com/cakes/",
  "provider": { "@id": "https://example.com/#identity" }
}

Notice provider points back to your identity node by its @id rather than repeating your name — more on that below.

The per-page nodes

On a single post or page, Agentimus adds nodes that describe that page.

The page itself

The main node describes the content and its type is chosen to fit:

  • A blog post becomes a BlogPosting.
  • A page becomes a WebPage.
  • Any other covered content type becomes a generic Article.
{
  "@type": "BlogPosting",
  "@id": "https://example.com/sourdough-guide/#blogposting",
  "url": "https://example.com/sourdough-guide/",
  "headline": "A Beginner's Guide to Sourdough",
  "datePublished": "2026-05-01T09:00:00+00:00",
  "dateModified": "2026-06-12T14:30:00+00:00",
  "author": { "@id": "https://example.com/#identity" },
  "publisher": { "@id": "https://example.com/#identity" },
  "mainEntityOfPage": "https://example.com/sourdough-guide/",
  "keywords": ["Sourdough", "Bread baking"],
  "about": [
    {
      "@type": "DefinedTerm",
      "name": "Sourdough",
      "sameAs": "https://www.wikidata.org/wiki/Q191600"
    }
  ]
}

The author and publisher both point back to your identity node, so your credit travels with the page. The keywords and about fields appear only when you have given the page topics — see the Topics for AI page for how those work, including the authoritative sameAs links on each topic.

A BreadcrumbList gives assistants the page’s place in your site — Home, then its first category, then the page itself. It helps them describe where a page sits.

{
  "@type": "BreadcrumbList",
  "itemListElement": [
    { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://example.com/" },
    { "@type": "ListItem", "position": 2, "name": "Recipes", "item": "https://example.com/category/recipes/" },
    { "@type": "ListItem", "position": 3, "name": "A Beginner's Guide to Sourdough", "item": "https://example.com/sourdough-guide/" }
  ]
}

FAQ pages (auto-detected)

If a page genuinely reads as a list of questions and answers, Agentimus emits a FAQPage node so assistants can lift your Q&A cleanly. You do not tag anything — it is detected from the content, using two deliberate, low-false-positive signals:

  1. Details/summary blocks — the native WordPress “Details” block, where the summary is the question and the body is the answer (a common accordion/FAQ pattern).
  2. Question headings — an h2, h3 or h4 whose text ends in a question mark, with the text that follows it (up to the next heading) as the answer.

To avoid spammy or accidental FAQ markup, Agentimus only publishes the node when it finds at least two genuine question-and-answer pairs — a single question is just prose. Answers are trimmed to a sensible length to keep the data tidy.

{
  "@type": "FAQPage",
  "@id": "https://example.com/sourdough-guide/#faq",
  "url": "https://example.com/sourdough-guide/",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "How long does a starter take to activate?",
      "acceptedAnswer": { "@type": "Answer", "text": "About five to seven days of daily feeding." }
    },
    {
      "@type": "Question",
      "name": "Can I use whole-wheat flour?",
      "acceptedAnswer": { "@type": "Answer", "text": "Yes — it actually ferments faster." }
    }
  ]
}

If you want a page to be recognised as an FAQ, write it that way: use the Details block, or use question headings that end in “?”, with the answer directly below.

Rather than repeating your name and details inside every node, Agentimus gives each site-level node a stable @id — a permanent anchor — and links to it from everywhere else:

  • Your WebSite is #website.
  • Your identity is #identity.
  • Each post’s node, breadcrumb and FAQ carry their own page-based @id (for example …/sourdough-guide/#blogposting and …#faq).

So a post’s author, its publisher, and every Service’s provider are all written as { "@id": "https://example.com/#identity" }. This tells a machine “the author here is the same entity defined above” — one authoritative description, referenced everywhere, with no risk of the details drifting apart. It is exactly how well-formed schema graphs are supposed to fit together, and it is what lets an assistant confidently connect an article to the person who wrote it.

When Agentimus stands down for an SEO plugin

Duplicate or conflicting structured data is one of the fastest ways to confuse a search engine — and one of the most common reasons schema gets ignored. So Agentimus checks whether a major SEO plugin is already emitting schema, and if so it stays completely silent on structured data (all its other features keep working).

It stands down when any of these is active:

  • Yoast SEO
  • Rank Math
  • SEOPress
  • The SEO Framework
  • All in One SEO

You do not have to choose — this is automatic. If you run one of these plugins, that plugin owns your schema and Agentimus won’t add a second, competing copy. The Readiness panel will show this as a passing check (“An SEO plugin owns schema; Agentimus is standing down to avoid duplicates”), so you can confirm at a glance that your site still has structured data, just from the other plugin.

The privacy guard: drafts and password-protected content

Structured data describes your public site, so Agentimus is careful never to leak content that isn’t public:

  • The site-level identity nodes (WebSite, your entity, Services) appear on public front-end pages as normal — they describe the site, not private content.
  • A per-page node is only emitted for content that is genuinely public. The post must be published (drafts, pending, scheduled and private posts get no per-page node) and it must not be password-protected.

The password rule is strict for a reason: a password-protected page’s body could contain question-and-answer text that would otherwise flow into the public FAQ node. Agentimus refuses to describe that content at all, so nothing behind a password ever surfaces in the structured data — even after the page is later published.

Turning it on, off, and previewing it

Structured data is controlled by the JSON-LD structured data toggle under Settings → Features. It is on by default. Reasons you might turn it off:

  • You run one of the SEO plugins above (though Agentimus already stands down automatically, so you can safely leave it on).
  • You have another dedicated schema plugin you prefer.

To see exactly what Agentimus would print for a given page before it ships, use the built-in Agent preview (available from the Readiness panel and from the editor). It shows the full JSON-LD graph — including a draft post’s would-be node, so you can check your markup before you publish. (The preview still respects the password guard: a password-protected page’s body is never shown, even in preview.)

Shaping what it says

Almost everything in the identity and Service nodes is yours to control, and it all lives in one place: the Identity page. That is where you set whether you are a Person or an Organization, your name and role, your “about” and “what this is not” descriptions, your audience, your areas of expertise, your verified sameAs profiles, and your services. Filling that page in thoroughly is the single most valuable thing you can do for your structured data — it turns an anonymous graph into a confident, verifiable statement of who you are.

For the per-page keywords and about topic data, see the Topics for AI page.

Back to top

Built and maintained by Sheikh Heera. Agentimus is free software (GPL-2.0-or-later); this documentation is generated from the plugin source — if something here disagrees with the code, the code wins.

This site uses Just the Docs, a documentation theme for Jekyll.