agentleFS
textura-agencytextura-agency/next16-claude-starter/.claude/skills/schema-markup/SKILL.md

schema-markup

Skill141 starsChanged 34 days agoUnlicenseView on GitHub

Generate and wire valid JSON-LD structured data for pages in this starter — Organization, WebSite, LocalBusiness, Article, Product, FAQPage, BreadcrumbList, Service — using the existing structured-data helper rather than inline script tags. Use when the user asks for "schema", "structured data", "JSON-LD", "rich snippets", "FAQ schema", or wants richer search/AI results.

68 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use keyword-cannibalization, broken-links and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • keuperaSkill

    keyword-cannibalization

    Detect and fix cases where several pages on the same site compete for the same query, splitting signals and suppressing all of them. Use whenever the user mentions keyword cannibalization, pages competing with each other, the wrong page ranking for a query, rankings that keep swapping between URLs, duplicate topics, or has published many posts on similar subjects. Also use when a page won't rank despite being good, or when Search Console shows several URLs alternating for one term.

    Also in 3 projects that use this

  • akii-technologies-ltdSkill

    broken-links

    Find and fix broken links — three modes with different scale ceilings. `local` audits any local repo (HTML/MD/MDX, no size limit) via Glob/Grep + per-link verification. `page` checks one live URL deeply (~30-100 outbound links per run). `site` requires the Ahrefs site-audit MCP for true multi-page crawl — without it, refuses and tells the user how to get there. Use when the user asks to "check broken links", "find dead links", "fix 404s", "link checker", "broken link audit", "redirect chains", "find dead URLs", "link validation", "mixed content". Accepts a mode modifier (`--mode=local|page|site`); auto-detects from the target if unspecified.

    Also in 3 projects that use this

  • JakeLabateSkill

    internal-link-audit

    Audit a website's internal linking structure for SEO. Crawls pages to find broken internal links, orphan pages, redirect chains, deep pages, missing or weak anchor text, and internal link distribution issues. Use when the user asks to audit, analyze, or improve internal links, site structure, link equity, or crawlability of a website.

    Also in 3 projects that use this

  • tifetolaSkill

    on-page-seo

    Also in 3 projects that use this

  • danielimadSkill

    technical-seo-audit

    Analyses a website crawl export (Screaming Frog, Sitebulb, or Ahrefs Site Audit) and returns a prioritised list of technical SEO issues. Finds redirect chains and loops, duplicate title tags and meta descriptions, pages Google cannot index or render, broken internal links, and orphan pages. Trigger when the user uploads or points to a crawl file, a CSV or XLSX of URLs with status codes, or asks for a technical SEO audit, a crawl analysis, or to "find redirect chains".

    Also in 4 projects that use this

  • xfstudioSkill

    product-marketing-context

    When the user wants to create or update their product marketing context document. Also use when the user mentions 'product context,' 'marketing context,' 'set up context,' 'positioning,' or wants to avoid repeating foundational information across marketing tasks. Creates `.claude/product-marketing-context.md` that other marketing skills reference.

    Also in 19 projects that use this

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 67 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use and extend those: Never hand-write a <script type="application/ld+json"> in a component.

    line 18
  2. Very specific:

    Check Google's current supported-features gallery (developers.google.com/search/docs/appearance/structured-data/search-gallery) before …

    line 24

“Common” is measured: kept by at least 6 in 10 of the 67 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Structured data (JSON-LD)

This project already has src/utils/seo/structured-data.ts: typed builders that each return a JsonLdNode — getSiteStructuredData() (Organization or a LocalBusiness subtype + WebSite, rendered once by the root layout), getBreadcrumbStructuredData(path) (from the route registry's parent chain), getArticleStructuredData, getServiceStructuredData, getFaqStructuredData, getPersonStructuredData, getLocalBusinessStructuredData (one location of a multi-location business) — plus jsonLdGraph(...nodes) and <JsonLd data> (src/components/common/json-ld.tsx, escapes <). Use and extend those. Never hand-write a <script type="application/ld+json"> in a component.

First: what is the markup for?

Schema.org validity does not make a page eligible for a Google rich result. Check Google's current supported-features gallery (developers.google.com/search/docs/appearance/structured-data/search-gallery) before promising one:

  • FAQ rich results stopped showing on 7 May 2026 (docs removed in June).
  • HowTo rich results were retired in 2023.
  • Sitelinks search box (WebSite + SearchAction) was retired in 2024; WebSite still feeds the site name.

Those types may still describe matching visible content — answer engines (ChatGPT, Perplexity, Claude) read them — but say so, and never sell them as a rich-result win. A missing report in Search Console for a retired feature is not a broken implementation.

Still worth it for Google: Organization (logo, knowledge panel facts), BreadcrumbList, Article/BlogPosting, Product + Offer, LocalBusiness, Event, Review/AggregateRating (on the reviewed thing only), VideoObject, JobPosting.

Rules

  1. JSON-LD only — never microdata or RDFa (hard rule #10).
  2. Only describe what is on the page. Schema that claims content the page does not contain is spam, and Google treats it that way.
  3. Reference, don't duplicate. Build the graph with @id references between entities so the Organization is defined once and linked from everywhere.
  4. Values come from siteConfig (src/lib/site.ts — organization holds the founding date, contact, address, sameAs, and the local facts openingHours, geo, areaServed, priceRange, hasMap), the route registry (src/lib/routes.ts) and the page's own props — never hardcoded a second time. Leave a fact out rather than guess one.
  5. Validate: the JSON parses, required properties are present, URLs are absolute.
  6. No placeholder names. The graph reads siteConfig.name — on sites built from the old starter the JSON-LD Organization said "New Project" while Lighthouse still scored SEO 100. logo points at the brand mark (/icon-512.png from the brand kit, or the logo file), never the starter's icon; sameAs lists only profiles the site actually links.
  7. The robot form carries the same graph. Crawlers get /robot-view under / (optimize-load/references/robot-path.md); render JSON-LD from the root layout or the view so both forms emit it, and diff the two: curl -s -A "Googlebot/2.1" <url>/ | grep -c 'application/ld+json' equals the people count.
  8. Absolute URLs come from siteConfig.url, which is the build-time origin — check the deployed graph names the real domain, not localhost.

Types worth adding, by page kind

PageTypeRequired-ish properties
HomeOrganization, WebSitename, url, logo, sameAs
Local businessLocalBusiness subtype — getSiteStructuredData({ type: "Dentist" })address, geo, openingHours, telephone (from siteConfig.organization; hours/geo/priceRange emit only for a LocalBusiness type)
One location of manygetLocalBusinessStructuredData({ path, name, address, … })its own @id (#location), parentOrganization → the brand — skill local-seo
Blog postArticle / BlogPostingheadline, image, datePublished, dateModified, author
Service pageServiceserviceType, provider, areaServed
ProductProduct + Offername, image, description, price, availability
FAQ blockFAQPagemainEntity[] of Question/acceptedAnswer — answer engines only; no Google rich result since May 2026
Any nested pageBreadcrumbListitemListElement with position
Team pagePerson per membername, jobTitle, image, worksFor

FAQPage no longer earns a Google rich result, but a real, visible FAQ is still one of the formats answer engines extract most reliably — mark it up with getFaqStructuredData when the questions are on the page, never schema-only. The answer text must match the page exactly.

Pattern

The site graph renders once in the root layout. A page's own graph renders in its view — one <JsonLd> per page:

const breadcrumb = getBreadcrumbStructuredData(PATH);
// …
<JsonLd data={jsonLdGraph(
  ...(breadcrumb ? [breadcrumb] : []),
  getServiceStructuredData({ path: PATH, name, description, serviceType }),
)} />

src/views/legal/privacy-policy.tsx ships the breadcrumb example. A type the helper lacks → add a typed builder to structured-data.ts (input interface, @id = ${absoluteUrl(path)}#<kind>, { "@id": ORGANIZATION_ID } for the publisher/provider), then document it in obsidian/frontend/utils.md.

yarn qa:seo FAILs on JSON-LD that doesn't parse and on a robot form whose block count differs from the people form.

After

Note new schema in obsidian/frontend/seo-metadata.md, and validate with Google's Rich Results Test / schema.org validator against the deployed URL. (When auditing another site: curl and WebFetch can miss JSON-LD a CMS plugin injects with JavaScript — check in a rendered browser before reporting "no schema".)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.