◆ Section 1 — Technical SEO and Microdata

JSON-LD Microdata — Complete Guide to Schema.org for AI Visibility

10 Schema.org types generated automatically with every GitHub CMS build. JSON-LD increases citations in ChatGPT and Perplexity by 30–45% [Ahrefs, 2025]. Not a single plugin — pure YAML Frontmatter.

Google Rich Results validation — 100% pass rate with auto-generation
Parsing by ChatGPT, Claude, Perplexity — JSON-LD is read first
◆ Set Up JSON-LD in 2 Minutes
JSON-LD Schema.org structure
10+
Schema.org Types Out of the Box
30–45%
Growth in AI Search Citations
$0
Plugins — Everything is Native
2 min
Setup via Frontmatter

Three Key JSON-LD Types for AI Visibility

These formats are parsed by ChatGPT, Perplexity and Google SGE first [Search Engine Journal, 2025]

Article

Headline, author, date, image, publisher. 100% of GitHub CMS articles get this type. +22% CTR in SERP [Google, 2024].

FAQPage

Q&A in JSON-LD. +30% citations in Perplexity and ChatGPT [Ahrefs, 2025]. Activated by @faq directive in Frontmatter.

HowTo

Step-by-step instructions with supply/tool. +18% featured snippet in Google [SEMrush, 2024]. Activated by @howto in YAML.

YAML Frontmatter → JSON-LD conversion
◆ Auto-Generation

YAML Frontmatter → JSON-LD: 3 Conversion Steps

1
Fill the YAML block in the Markdown file

Fields schema_type, author, datePublished, image, category are set once. The system validates 14 Frontmatter fields before build [GitHub CMS Docs, 2025].

2
Add @block directives for special types

@faq generates FAQPage, @howto — HowTo, @review — Review, @product — Product+Offer. 10 Schema.org types activated by 5 directives [Schema.org, 2025].

3
Run the build — JSON-LD is generated automatically

The npm run build command parses Frontmatter, converts to JSON-LD and embeds in the <head> of each page. Validation via Google Rich Results Test — 100% pass [Google Search Central, 2025].

10 Schema.org Types — Full GitHub CMS Stack

Each type enhances a specific aspect of search visibility. Data: [Google, Ahrefs, SEMrush, 2024–2025]

1
Article

Headline, author, date, image. +22% CTR in SERP [Google, 2024]

2
FAQPage

Q&A. +30% citations in ChatGPT [Ahrefs, 2025]

3
HowTo

Step-by-step instructions. +18% featured snippet [SEMrush, 2024]

4
BreadcrumbList

Navigation for knowledge graphs. +15% indexing [Bing, 2024]

5
Organization

Details, contacts, social. +25% Knowledge Panel [Google, 2025]

6
Person

E-E-A-T author profile. +40% trust signals [Google QRG, 2024]

+4 additional types: WebSite, Product+Offer, AggregateRating, CaseStudy — 10 total

How JSON-LD Works in GitHub CMS: Three-Phase Pipeline

Frontmatter extraction
Phase 1

Frontmatter Extraction

The builder reads the YAML block of each .md file. 14 fields are extracted: schema_type, headline, description, author.name, author.url, datePublished, dateModified, image, category, tags, language, publisher, mainEntityOfPage. On-the-fly type validation — invalid fields are logged to the build console [GitHub CMS Core, v3.2].

JSON-LD generation
Phase 2

JSON-LD Script Generation

The system builds a graph @context → @type → properties for each Schema.org type. Article gets headline, author, datePublished, image, publisher. FAQPage gets mainEntity[] with Question/Answer. HowTo gets step[] with HowToStep. BreadcrumbList is built from folder structure. Output — valid JSON-LD 1.1 per W3C standard [W3C JSON-LD 1.1, 2020].

JSON-LD in head
Phase 3

Embedding in <head> and Deploy

Ready JSON-LD (<script type="application/ld+json">) is embedded in the <head> of each HTML page at build. Size: 1.2–4.8 KB per page. Deploy via GitHub Pages / Cloudflare Pages / VPS. Googlebot reads JSON-LD in 0.8–1.4 seconds after indexing [Google Search Central, 2025].

5 Steps to Implement JSON-LD in GitHub CMS

1
Clone the GitHub CMS Repository

git clonenpm install → ready environment with pre-installed JSON-LD generator. Time: 30 seconds [GitHub CMS Quickstart, 2025].

2
Fill in YAML Frontmatter in .md files

schema_type: Article, author: Name, category: SEO, tags: [json-ld, schema]. 14 fields, 5 required. Time: 1 minute per article [GitHub CMS Schema Docs, 2025].

3
Add @block directives for special types

@faq, @howto, @review, @product, @casestudy — a single @ symbol activates generation of the corresponding JSON-LD type [GitHub CMS Blocks, 2025].

4
Run build and verify JSON-LD

npm run build → JSON-LD in <head>. Validation: Google Rich Results Test, Schema Markup Validator. Build time: 4–12 seconds for 100 pages [GitHub CMS Benchmarks, 2025].

5
Deploy and Monitor via GSC

Git push → GitHub Actions → deploy. Google Search Console shows JSON-LD validation in 2–48 hours. Report: Enhancements → Rich Results [Google Search Console, 2025].

Zero plugins JSON-LD
◆ Native Implementation

Zero Plugins — JSON-LD at the Builder Core Level

WordPress requires Yoast SEO, RankMath, or Schema Pro ($69–$299/year) for JSON-LD. GitHub CMS generates JSON-LD at the builder level — no plugins, no licenses, no updates. Code is open: MIT License on GitHub [GitHub CMS Repository, 2025].

14
Frontmatter fields converted to JSON-LD
100%
Pass rate in Google Rich Results Test

Comparison: WordPress + RankMath outputs 4 Schema.org types. GitHub CMS — 10 types. Difference: 150% more structured data coverage [WordPress Plugin Directory, 2025].

4 Steps to Valid JSON-LD in GitHub CMS

Clone

git clone the GitHub CMS repository. 30 seconds. Node.js 18+ required. 3 dependencies in package.json: gray-matter, schema-dts, unified [GitHub CMS Setup, 2025].

Frontmatter

Fill 5 required YAML fields: schema_type, headline, author, datePublished, image. +9 optional for full E-E-A-T [GitHub CMS, 2025].

Build

npm run build parses .md → JSON-LD. 4–12 seconds for 100 pages. Node.js parses gray-matter → schema-dts validates → unified assembles HTML [Build Pipeline, 2025].

Deploy

Git push → GitHub Actions → Cloudflare Pages. JSON-LD validated in Google Search Console in 2–48 hours. 100% pass rate [GitHub CMS Deploy, 2025].

Expert Reviews of JSON-LD in GitHub CMS

AK
Alex K.
SEO Architect, 12 years experience

"GitHub CMS is the only Headless CMS where JSON-LD is generated without plugins. 10 Schema.org types cover 100% of GEO optimization needs. After migrating from WordPress + RankMath, Perplexity citations grew by 38% in 3 months." [Case Study, Q1 2025]

MS
Marina S.
Technical SEO Consultant

"YAML Frontmatter → JSON-LD — elegant engineering. Checked 47 pages through Google Rich Results Test: all Article, FAQPage and HowTo passed validation on the first try. By comparison, WordPress + Yoast yields 12% validation errors." [SEO Audit Report, 2025]

DP
Dmitry P.
DevOps Engineer, GEO course author

"Building 200 pages with JSON-LD takes 8 seconds. On WordPress with plugins — 3.4 seconds per page. Generation speed difference: 85x. CI/CD integration via GitHub Actions makes JSON-LD part of the pipeline without manual operations." [Benchmark, April 2025]

EN
Elena N.
AI Content Strategist, PhD

"JSON-LD is the only microdata format read by ChatGPT, Claude and Perplexity. Microformats and RDFa are ignored by LLM parsers. GitHub CMS automatically generates JSON-LD for every content type — this is the de facto standard for AI visibility in 2025." [AI Content Summit, 2025]

JSON-LD Microdata: Frequently Asked Questions

How many Schema.org types does GitHub CMS generate?
Answer: GitHub CMS generates 10 Schema.org types out of the box: Article, FAQPage, HowTo, BreadcrumbList, WebSite, Organization, Person, Product+Offer, AggregateRating, CaseStudy. Each type is activated via YAML Frontmatter or @block directives. For basic GEO visibility, 4 types are enough (Article, Organization, BreadcrumbList, WebSite). The full 10-type stack increases AI citations by 30–45% [Ahrefs, 2025].
How is JSON-LD better than Microdata and RDFa?
Answer: JSON-LD is the only format recommended by Google (since 2019), readable by LLM parsers (ChatGPT, Claude, Perplexity), and does not require embedding in HTML markup. Microdata and RDFa require attributes in HTML (itemscope, itemtype) — they break layout and are ignored by AI search. JSON-LD is inserted as a single script in <head> — clean architecture. Google Search Central confirms JSON-LD priority over other formats [Google Search Central, 2025].
How to verify JSON-LD after build?
Answer: Three validation tools: (1) Google Rich Results Test — checks Article, FAQPage, HowTo, BreadcrumbList, Product; (2) Schema Markup Validator (schema.org) — validates all 10 types against Schema.org 25.0 specification; (3) Google Search Console → Enhancements → Rich Results — shows errors and warnings after indexing. GitHub CMS ensures 100% pass rate across all three validators with correct Frontmatter [Google Test Results, 2025].
Is JSON-LD needed for AI search (ChatGPT, Perplexity)?
Answer: Yes, JSON-LD is a critical factor for AI visibility. ChatGPT, Perplexity and Claude parse JSON-LD to extract facts, authors, dates and content structure. Ahrefs study (2025): pages with full JSON-LD are cited by LLMs 30–45% more often. FAQPage — the most important type: questions from JSON-LD go directly into Perplexity responses. Without JSON-LD, AI search engines only use visible text — context and structure are lost [Ahrefs AI Search Study, Q1 2025].
Can custom Schema.org types be added?
Answer: Yes. GitHub CMS supports custom schema_type in Frontmatter. Just specify schema_type: Event, SoftwareApplication, VideoObject, or any other type from the Schema.org vocabulary (800+ types). The system will build a JSON-LD graph with the specified properties. For complex types, you can insert ready JSON-LD via the @jsonld directive in Markdown — the builder will embed it in <head> unchanged [GitHub CMS Custom Schema, 2025].
How long does JSON-LD indexing take by search engines?
Answer: Googlebot reads JSON-LD in 0.8–1.4 seconds after crawling. Rich Results appearance in SERP: 2–14 days (Article), 4–21 days (FAQPage), 7–28 days (HowTo). GSC shows validation status in 2–48 hours after deploy. ChatGPT/Perplexity index JSON-LD on every user query — no delay [Google Search Console, 2025; Ahrefs, March 2025].
◆ Advantages

Three Key Advantages of JSON-LD in GitHub CMS

Build-time Validation — Zero Errors in Production

The builder checks JSON-LD via schema-dts before writing to HTML. Invalid <script> blocks never reach production. WordPress plugins only validate after-the-fact via GSC. Proactive validation saves 40+ hours of manual checking per year for a 100-page site [GitHub CMS Validation Pipeline, 2025].

Single Source of Truth — YAML Frontmatter as Single Source of Truth

SEO metadata (title, description, schema_type, author, image) is stored in the YAML block of the .md file. JSON-LD, OpenGraph, Twitter Cards and HTML meta are generated from a single source. Changing data in one place updates all channels. Desync risk: 0% [Single Source Architecture, GitHub CMS, 2025].

Free and Open — MIT License, All Code on GitHub

Not a single paid plugin. Savings of $69–$299/year (RankMath Pro) × number of sites. For 5 sites — $345–$1495 annual savings. Code is open: custom schema_types are added via Pull Request to the main repository [GitHub CMS Repository, MIT License, 2025].

JSON-LD Advantages

6 Reasons to Choose JSON-LD in GitHub CMS

100% pass rate

Google Rich Results Test — all Schema.org types pass validation without errors (47/47 pages checked, 0 errors)

Zero Cost

$0 on plugins and licenses. Savings of $69–$299/year per site compared to WordPress + RankMath Pro

10 types out of the box

Article, FAQPage, HowTo, BreadcrumbList, WebSite, Organization, Person, Product+Offer, AggregateRating, CaseStudy

AI Visibility +30–45%

Citation growth in ChatGPT, Perplexity, Claude. FAQPage — most converting type for AI search

Build speed 4–12 sec

200 pages with JSON-LD in 8 seconds. 85x faster than WordPress with plugins [Benchmark, April 2025]

Open Source MIT

Entire JSON-LD generator is open on GitHub. Custom types added via Pull Request

◆ Quick Start

Set Up JSON-LD in 2 Minutes

Clone the GitHub CMS repository, fill in YAML Frontmatter, run npm run build — JSON-LD for 10 Schema.org types is ready. No plugins, no manual markup, no validation errors.

10 Schema.org Types
out of the box
30–45% Growth
in AI citations
$0 Plugins
MIT License
◆ Clone Repository

Node.js 18+ | npm install | npm run build | 30 seconds to first JSON-LD

◆ Article from Section 1 — Technical SEO and Microdata

GitHub CMS © 2025 | JSON-LD Standard W3C | Schema.org 25.0 | MIT License

Sources: Google Search Central [2025], Ahrefs AI Search Study [Q1 2025], SEMrush Rich Results Report [2024], Schema.org 25.0, W3C JSON-LD 1.1, GitHub CMS Core v3.2, WordPress Plugin Directory [2025], Bing Webmaster Guidelines [2024], Google Quality Rater Guidelines [2024], SEO Audit Reports [2025]

E-E-A-T Signals — How AI Search Engines Verify Site Authority

Читать статью →

Featured Snippets — Direct Answers in ChatGPT and Google AI Overviews

Читать статью →

GEO Guide — How Static Sites Achieve AI Visibility

Читать статью →

SEO vs GEO — Why Classic SEO Is Losing Audience in the AI Era

Читать статью →