Claude Code Skills for Schema Markup: 9 Free and Paid Options (2026)

Google turned off FAQ rich results on May 7, 2026. If your schema skill still offers to “add FAQ dropdowns to your SERP listing,” it is working from an old rulebook.

Quick pick: the strongest free options are seo-schema inside AgriciDaniel’s claude-seo plugin, the schema skill in Corey Haines’ marketingskills, and Alireza Rezvani’s schema-markup skill, which ships a Python validator. All three are MIT licensed. Whatever you install, check its output in Google’s Rich Results Test before shipping.

That one date is the best filter for picking a schema skill this year. Some popular skills already know about it. Others still mention FAQ dropdowns and HowTo steps as rich result targets. This guide covers which skills detect, validate and generate JSON-LD correctly. It also covers which schema types still earn rich results, how to check output with the Rich Results Test, and real JSON-LD you can copy.

Every star count, push date and price below was checked on the live page on 1 November 2026.

What is a Claude Code skill for schema markup?

A Claude Code skill for schema markup is a folder with a SKILL.md file that teaches Claude how to find, check and write structured data (usually JSON-LD) for a page. Claude loads it when you ask about schema, structured data or rich results. The good ones add scripts that parse HTML and validate required properties, and reference files that track Google’s current feature list.

If you are new to the format, read what Claude Code skills are first. The short version:

• A skill lives in ~/.claude/skills/<name>/ (all projects) or .claude/skills/<name>/ (one project).
• You can install it from a GitHub repo with npx skills add owner/repo, or as a plugin through /plugin marketplace add and /plugin install.
• Plugins can bundle skills, subagents, hooks and MCP server configs together.

The full walkthrough is in how to install Claude Code skills.

The three jobs a schema skill has to do

Job	What it means	What a weak skill does instead
Detect	Pull every application/ld+json block, plus Microdata and RDFa, from a URL or file	Reads only what you paste
Validate	Check required and recommended properties against Google’s current docs	Checks JSON syntax and stops
Generate	Write JSON-LD that matches visible page content and a type Google still supports	Writes FAQPage and HowTo for everything

A skill that only generates is a template library. Templates are useful, but they will not stop you shipping markup that no longer earns anything in Google.

Takeaway: judge a schema skill by its validation and its deprecation list, not by how many types it can write.

Which schema types still produce rich results in 2026?

As of November 2026, Google’s search gallery lists 25 structured data features, including Article, Breadcrumb, Product, Review snippet, Local business, Event, Recipe, Video, Job posting, Organization, Profile page, Discussion forum, Q&A and Software app. FAQ and HowTo are no longer on it.

Here is the timeline, taken from Google’s own Search Central changelog:

Date	Change
Sept 2023	HowTo rich results removed from Search
Aug 2023	FAQ rich results limited to authoritative government and health sites
23 Apr 2025	Deprecation notice added for SpecialAnnouncement (feature ends 31 July 2025)
12 Jun 2025	Deprecation banners on book actions, course info, estimated salary, ClaimReview, learning video, special announcement, vehicle listing
9 Sep 2025	Docs removed for course info, estimated salary, learning video, special announcement, vehicle listing
5 Nov 2025	Practice problem deprecation notice; book actions banner removed because a feature still uses it
6 Jan 2026	Practice problem docs removed
7 May 2026	FAQ rich results stop appearing in Google Search for all sites
15 Jun 2026	FAQ rich result documentation removed

Types worth marking up for Google right now

Type	Rich result it supports	Typical page
Article / BlogPosting	Article display in Top stories and Discover surfaces	Blog posts, news
BreadcrumbList	Breadcrumb trail in place of the URL	Any page deeper than the homepage
Product + Offer + Review	Price, availability, review stars, merchant listings	Product pages
LocalBusiness	Business details tied to the knowledge panel	Location pages
Organization	Logo and company details	Homepage, about page
Event	Event listings	Ticketed or dated events
Recipe	Recipe cards and carousels	Recipe posts
VideoObject	Video previews and key moments	Pages with an embedded video
JobPosting	Google for Jobs listings	Careers pages
SoftwareApplication	App rating and price	SaaS and app pages
DiscussionForumPosting / QAPage	Forum and Q&A features	Community threads
ProfilePage	Creator profile details	Author and creator pages

Should you delete your FAQPage markup?

No. Google’s guidance on retired features is that you can drop the markup, but there is no need to remove it proactively, because unused structured data does not cause problems in Search. FAQPage is still valid schema.org vocabulary. Bing and other consumers may still read it.

What you should stop doing is adding FAQPage to get a Google rich result. Many SEO blogs now claim FAQ schema helps ChatGPT or Perplexity cite you. That claim is not confirmed by any of those engines, so treat it as a hypothesis, not a reason.

This is also why we left FAQPage out of this post’s own schema field above.

Takeaway: mark up the types on Google’s current gallery, keep old FAQPage if it matches visible content, and stop selling FAQ schema as a SERP feature.

The best free Claude Code skills for schema markup

The strongest free options are AgriciDaniel’s seo-schema (inside the claude-seo plugin), Corey Haines’ schema skill in marketingskills, and Alireza Rezvani’s schema-markup skill, which ships a Python validator. All three are MIT licensed and were pushed to within the last two days of our check.

We read each SKILL.md on GitHub. Here is what they actually contain.

1. claude-seo: seo-schema (AgriciDaniel)

• Source: github.com/AgriciDaniel/claude-seo, skill at skills/seo-schema/
• Price: Free (MIT)
• Stars / last push: 25,381 stars, pushed 1 Nov 2026
• What it does: Detection, validation and generation in one skill. It extracts JSON-LD with a bundled parse_html.py script, checks for Microdata and RDFa, and flags blocks missing @context or @type. Its SKILL.md has a “Schema Type Status (checked October 2026)” section that marks FAQPage as retired on 7 May 2026 and lists HowTo, SpecialAnnouncement and Practice Problem as never-recommend. A reference file, deprecated-types-2024-2026.md, holds the full list. The repo also ships hooks/validate-schema.py and scripts/schema_generate.py.
• Install:

/plugin marketplace add AgriciDaniel/claude-seo
/plugin install claude-seo@agricidaniel-claude-seo

Directory listings also show npx skills add agricidaniel/claude-seo --skill seo-schema for the single skill. Usage: /seo-schema https://example.com/page.

Best for: technical SEOs who want schema checks inside a wider audit. It pairs with the rest of the suite covered in Claude Code skills for SEO audits.

2. marketingskills: schema (Corey Haines)

• Source: github.com/coreyhaines31/marketingskills, skill at skills/schema/
• Price: Free (MIT)
• Stars / last push: 57,803 stars, pushed 31 Oct 2026
• What it does: Starts with an assessment (page type, current schema, goals), then writes JSON-LD from a schema-examples.md reference. It reads your product-marketing.md context file if one exists, so brand names and URLs come out consistent. Version 2.0.2 states plainly that Google stopped FAQ rich results on 7 May 2026 and retired HowTo in 2023, and tells Claude not to recommend those types for a retired Google feature. It also ships an evals/evals.json file, which means the author tests the skill’s behaviour.
• Install:

npx skills add coreyhaines31/marketingskills --skill schema

Best for: marketers who already use the rest of the pack. See Claude Code skills for marketers.

3. claude-skills: schema-markup (Alireza Rezvani)

• Source: github.com/alirezarezvani/claude-skills, skill at marketing-skill/skills/schema-markup/
• Price: Free (MIT)
• Stars / last push: 29,110 stars, pushed 31 Oct 2026
• What it does: Runs scripts/schema_validator.py against page HTML to check Google’s required properties, then points you to the Rich Results Test and validator.schema.org. Two reference files cover implementation patterns and a type guide.
• Watch out: The SKILL.md metadata says updated: 2026-03-06, before the FAQ change. It still lists “FAQ dropdowns” and “HowTo steps” as rich result targets and maps how-to guides to HowTo. Use the validator, and ignore those two suggestions.
• Install:

npx skills add alirezarezvani/claude-skills --skill schema-markup

Or add the marketplace with /plugin marketplace add alirezarezvani/claude-skills and install the marketing plugin.

4. claude-code-templates: schema-markup (davila7)

• Source: github.com/davila7/claude-code-templates, at cli-tool/components/skills/business-marketing/schema-markup/
• Price: Free (MIT)
• Stars / last push: 33,036 stars, pushed 1 Nov 2026
• What it does: Covers Article, Product, FAQ, BreadcrumbList, LocalBusiness and HowTo patterns. Check how it treats FAQ and HowTo before you trust its suggestions.
• Install: through the project’s CLI, npx claude-code-templates@latest, then pick the skill.

5. marketing-skills: schema (kostja94)

• Source: github.com/kostja94/marketing-skills, at skills/seo/on-page/schema/
• Price: Free (MIT)
• Stars / last push: 1,071 stars, pushed 26 Oct 2026
• What it does: On-page schema guidance that also frames markup for AI search visibility.
• Install: npx skills add kostja94/marketing-skills (use --list to pick only the schema skill)

6. openclaudia-skills: schema-markup

• Source: github.com/OpenClaudia/openclaudia-skills, at skills/schema-markup/
• Price: Free (MIT)
• Stars / last push: 724 stars, pushed 29 Oct 2026
• Install: npx skills add OpenClaudia/openclaudia-skills --skill schema-markup

7. aaron-marketing-skills: serp-markup-builder (Aaron He Zhu)

• Source: github.com/aaron-he-zhu/aaron-marketing-skills
• Price: Free
• Stars / last push: 3,046 stars, pushed 1 Nov 2026
• What it does: The old schema-markup-generator from seo-geo-claude-skills was merged into serp-markup-builder. The old repo is now a signpost, so install the new one.
• Install: npx skills add aaron-he-zhu/aaron-marketing-skills -s serp-markup-builder

Takeaway: if you install one free skill, pick one whose SKILL.md already names the 7 May 2026 FAQ change. Two of the big three do.

Paid skills and SaaS tools for schema markup

Paid options fall into two groups: skill packs sold as Markdown files, and schema SaaS platforms that expose your data to Claude through an MCP server. Only one skill pack we found had a public price and no fake scarcity.

8. The Ultimate Claude SEO & AEO Skills Pack (Contra)

• Source: contra.com/products/fE9zVnna-the-ultimate-claude-seo-and-aeo-skills-pack, by Samuel Uebersax
• Price: $39 (shown against a struck-through $79). No countdown or “spots left” messaging on the page.
• What it does: 12 skill files in .md format, including a Schema Markup Generator (Article, Product, FAQ, LocalBusiness, HowTo, BreadcrumbList) and an SEO site auditor that covers structured data. Two bonus skills are Webflow-specific. Lifetime updates are listed.
• Install: download the files and drop each into ~/.claude/skills/<name>/SKILL.md.
• Caveat: the product page lists FAQ and HowTo among generated types. Check that the generator does not pitch them as Google rich results.

9. Schema App MCP server

• Source: schemaapp.com, endpoint https://mcp.schemaapp.com/mcp
• Price: Platform pricing not public (custom quote, 12-month minimum, onboarding fee). The pricing page lists the MCP server as free for internal team use, with external chatbot use at up to 100K requests per month included and $100 per extra 100K.
• What it does: Connects Claude to your Schema App Content Knowledge Graph. Schema App’s own FAQ says it is for your owned AI tools and does not improve AEO or GEO.
• Install: claude mcp add --transport http schemaapp https://mcp.schemaapp.com/mcp, then sign in with your Schema App account.

Also worth knowing: WordLift

WordLift builds knowledge graphs and schema markup for sites. Business+ is €799 per month billed yearly or €999 billed monthly, and Enterprise is custom. A WordLift MCP server appears on claudemarketplaces.com, but WordLift’s pricing page does not mention MCP, and we could not confirm setup details on WordLift’s own docs. Verify with them before you buy for this use.

General SEO MCP servers (DataForSEO, Ahrefs, Semrush, Google Search Console) are paid or need accounts, but they are not schema tools. Search Console is the one that matters here, because its rich result reports show what Google parsed. More on those in Claude Code skills for technical SEO.

Free vs paid: comparison table

Option	Free or paid	Price	Detect	Validate	Generate	Knows FAQ change (May 2026)	Install
claude-seo seo-schema	Free	$0 (MIT)	Yes, script	Yes, plus hook	Yes	Yes	/plugin install
marketingskills schema	Free	$0 (MIT)	Manual	Guidance	Yes	Yes	npx skills add
claude-skills schema-markup	Free	$0 (MIT)	Via script	Yes, Python	Yes	No (updated Mar 2026)	npx skills add
claude-code-templates	Free	$0 (MIT)	Manual	Guidance	Yes	Not confirmed	CLI
kostja94 schema	Free	$0 (MIT)	Manual	Guidance	Yes	Not confirmed	npx skills add
openclaudia schema-markup	Free	$0 (MIT)	Manual	Guidance	Yes	Not confirmed	npx skills add
serp-markup-builder	Free	$0	Manual	Guidance	Yes	Not confirmed	npx skills add
Contra SEO & AEO pack	Paid	$39	Via auditor	Not stated	Yes	Not confirmed	Manual copy
Schema App MCP	Paid platform	Pricing not public; MCP free for internal use	Reads your graph	Platform-side	Platform-side	n/a	claude mcp add

“Not confirmed” means we did not find the change named in the skill text we read. It does not mean the skill is wrong.

Takeaway: for most people, a free skill does everything a paid pack does for schema. Pay for a platform only when you need schema managed across thousands of URLs.

How to detect, validate and generate JSON-LD with Claude Code

The workflow is three passes: extract what exists, check it against Google’s current docs, then generate only what is missing and matches visible content. Here is a prompt sequence that works with any of the skills above.

Step 1: Detect

Extract every JSON-LD, Microdata and RDFa block from https://example.com/blog/post.
List each @type, flag blocks without @context or @type, and flag duplicates.

With claude-seo installed this calls its parse script. Without it, Claude can fetch the page and read the <script type="application/ld+json"> tags. Note that markup injected by JavaScript may not appear in raw HTML. Google renders JavaScript, so the Rich Results Test is the source of truth there.

Step 2: Validate

For each block, check required and recommended properties against Google's
current structured data docs. Mark any type Google no longer supports for
rich results, and say whether keeping it causes harm.

Then run two outside checks:

1. Rich Results Test: tells you which Google rich results the page is eligible for. It no longer reports on FAQ.
2. Schema Markup Validator: checks schema.org syntax for every type, including ones Google ignores.

Passing the Rich Results Test means eligible, not guaranteed. Google decides whether to show the feature.

Step 3: Generate

Write JSON-LD for this page using only types on Google's current search gallery.
Every value must match visible text on the page. Use @id to connect
Organization, WebSite, WebPage and Article into one @graph.

Then paste the output back into the Rich Results Test using the code snippet tab, before it ships.

Step 4: Monitor

After deploy, use the enhancement reports in Google Search Console. They show valid items, warnings and errors per type, which is the only place you see what Google actually parsed at scale.

Takeaway: never ship generated JSON-LD without a Rich Results Test pass and a human check that every value appears on the page.

Real JSON-LD examples you can copy

These three examples cover the most common jobs: a blog post, a product page and a local business. Replace the example.com values with your own. Every value must appear on the page.

Article with Organization and breadcrumbs

{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://example.com/#org",
      "name": "Example Co",
      "url": "https://example.com/",
      "logo": "https://example.com/logo.png"
    },
    {
      "@type": "BlogPosting",
      "@id": "https://example.com/blog/schema-guide/#article",
      "headline": "How to add schema markup to a blog post",
      "datePublished": "2026-10-20",
      "dateModified": "2026-11-01",
      "author": { "@type": "Person", "name": "Jane Doe", "url": "https://example.com/authors/jane-doe" },
      "publisher": { "@id": "https://example.com/#org" },
      "image": "https://example.com/images/schema-guide.png",
      "mainEntityOfPage": "https://example.com/blog/schema-guide/"
    },
    {
      "@type": "BreadcrumbList",
      "itemListElement": [
        { "@type": "ListItem", "position": 1, "name": "Blog", "item": "https://example.com/blog/" },
        { "@type": "ListItem", "position": 2, "name": "Schema guide", "item": "https://example.com/blog/schema-guide/" }
      ]
    }
  ]
}

Product with offer and rating

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Trail Runner 2 Shoe",
  "image": "https://example.com/img/trail-runner-2.jpg",
  "sku": "TR2-BLK-42",
  "brand": { "@type": "Brand", "name": "Example Co" },
  "offers": {
    "@type": "Offer",
    "url": "https://example.com/p/trail-runner-2",
    "price": "129.00",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock",
    "itemCondition": "https://schema.org/NewCondition"
  },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.4",
    "reviewCount": "87"
  }
}

Only include aggregateRating if those reviews are real and shown on the page. Self-serving review markup on your own business page (LocalBusiness or Organization reviewing itself) is not eligible for review stars.

LocalBusiness

{
  "@context": "https://schema.org",
  "@type": "Plumber",
  "name": "Example Plumbing",
  "url": "https://example.com/",
  "telephone": "+1-555-010-0199",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "100 Main St",
    "addressLocality": "Springfield",
    "addressRegion": "IL",
    "postalCode": "62701",
    "addressCountry": "US"
  },
  "openingHoursSpecification": [{
    "@type": "OpeningHoursSpecification",
    "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"],
    "opens": "08:00",
    "closes": "18:00"
  }]
}

Use the most specific LocalBusiness subtype that fits. For multi-location work see Claude Code skills for local SEO, and for templated pages at scale see Claude Code skills for programmatic SEO.

Common schema mistakes Claude Code skills catch (and some they cause)

The most common errors are markup that does not match visible content, missing required properties, retired types pitched as rich results, and duplicate blocks from plugins and themes. A good skill catches the first three. Only you can catch the fourth reliably, because it depends on your stack.

Mistakes a skill should flag

• Missing required properties. Product without offers, review or aggregateRating cannot get product snippets.
• Wrong value formats. Dates not in ISO 8601, prices with currency symbols inside price, availability as plain text instead of a schema.org URL.
• Orphan blocks. A block with no @context or @type cannot attach to an entity. claude-seo flags this.
• Retired types sold as SERP features. FAQ and HowTo are the big two.

Mistakes a skill can cause

• Inventing values. Ask for a review count and an eager model may fill one in. Every number in schema has to come from the page.
• Duplicates. Your SEO plugin, theme and new skill output may all write Organization. Detect first, then generate.
• Stale rules. A skill frozen in March 2026 does not know about May 2026. Check the skill’s last update and its deprecation section before you trust it.

If you want to fix a stale skill, editing the SKILL.md is simple. See how to build a Claude Code skill.

Takeaway: the riskiest output from any schema skill is a plausible number that is not on the page.

How to choose: a quick checklist

Pick the skill that matches how you work, check that it knows the 2026 FAQ change, and confirm it validates rather than only writes.

1. Open the SKILL.md on GitHub. Search it for “FAQ” and “HowTo”. Does it call them retired?
2. Check the last push date with gh api repos/OWNER/REPO --jq .pushed_at.
3. Look for scripts (parse_html.py, schema_validator.py). Scripts give repeatable checks.
4. Check the licence. All the free options above are MIT, except aaron-marketing-skills (check its repo).
5. Run it on one page and compare its output to the Rich Results Test.
6. If you also write content, combine it with a writing skill from Claude Code skills for content writers so schema ships with each post.

For the AI search side of structured data, see Claude Code skills for GEO. For the wider SEO stack, start at the best Claude Code skills for SEO.

FAQ

Does Google still show FAQ rich results in 2026?

No. Google stopped showing FAQ rich results for all sites on 7 May 2026 and removed the documentation on 15 June 2026. Before that, since August 2023, only authoritative government and health sites could get them.

Should I remove FAQPage schema from my site?

You do not need to. Google says unused structured data does not cause problems. Keep it if it matches visible content. Just stop adding it for a Google rich result.

Are HowTo rich results still available?

No. Google removed HowTo rich results in September 2023. HowTo is still valid schema.org vocabulary, but Google does not use it for a search feature.

Does schema markup help rankings?

Google has said structured data is not a direct ranking factor. It helps Google understand the page and makes it eligible for rich results, which can change how a listing looks.

Does schema markup help with ChatGPT, Perplexity or AI Overviews?

It is unproven. Many blogs claim it helps, but none of the AI engines has confirmed that schema affects citation. Structured, clear content on the page is the safer bet.

What is the best Claude Code skill for schema markup?

For audits, AgriciDaniel’s seo-schema, because it combines detection scripts with an October 2026 deprecation list. For marketers, Corey Haines’ schema skill. Both are free.

Can Claude Code generate JSON-LD without a skill?

Yes. Claude can write JSON-LD from a prompt. A skill adds consistent rules, current deprecation knowledge and validation scripts, so results do not depend on how well you phrase each prompt.

How do I validate schema markup?

Use Google’s Rich Results Test for Google eligibility and the Schema Markup Validator at validator.schema.org for schema.org syntax. After launch, watch the enhancement reports in Search Console.

Is JSON-LD better than Microdata?

Google recommends JSON-LD because it is easier to maintain. Microdata and RDFa still work. Most skills, including claude-seo, recommend JSON-LD by default.

Why does the Rich Results Test say my page is eligible but no rich result shows?

Eligibility is not a guarantee. Google decides whether to show a rich result based on quality, query and other factors.

Where do I put JSON-LD on the page?

In a <script type="application/ld+json"> tag, in the head or body. Google reads both, and it can read JSON-LD injected with JavaScript.

How do I install a schema skill in Claude Code?

Run npx skills add owner/repo --skill name, or add a plugin marketplace with /plugin marketplace add owner/repo and install with /plugin install. Restart Claude Code if the skill does not show.

Which schema types should every site have?

Organization (or LocalBusiness) on the homepage, BreadcrumbList on inner pages, and the type that matches each page’s main content, such as Article or Product.

Are paid schema skill packs worth it?

For schema alone, the free skills above cover detection, validation and generation. A paid pack makes sense if it bundles other skills you need. Platforms like Schema App suit large sites that need schema managed centrally.

Can a schema skill add fake reviews or ratings?

It can if you let it. Only mark up ratings and reviews that are real and visible on the page. Markup that misrepresents content can lead to a manual action.

Sources and methodology

We checked every repo with the GitHub API (gh api repos/OWNER/REPO) on 1 November 2026 for stars, last push and licence, and read each schema SKILL.md from the default branch. Prices were read from the live product and pricing pages on the same day.

• Google Search Central changelog: https://developers.google.com/search/updates
• Google search gallery: https://developers.google.com/search/docs/appearance/structured-data/search-gallery
• Google HowTo and FAQ changes (2023): https://developers.google.com/search/blog/2023/08/howto-faq-changes
• Rich Results Test: https://search.google.com/test/rich-results
• Schema Markup Validator: https://validator.schema.org
• Schema App pricing: https://www.schemaapp.com/pricing/
• WordLift pricing: https://wordlift.io/pricing/
• Contra pack: https://contra.com/products/fE9zVnna-the-ultimate-claude-seo-and-aeo-skills-pack

We did not run each skill against a live site for this post. Star counts change daily.

Your next step

Install one free skill that knows the May 2026 FAQ change, run it on your three most important page templates, and push each result through the Rich Results Test. Fix errors, delete nothing that matches the page, and stop adding markup for features Google no longer shows.