Section 3 · Content / Markdown
YAML Frontmatter — Complete Guide to 15 GitHub CMS Fields in 2 Minutes
title, description, slug, author, category, tags, schema_type, raw_html, and more. Each field directly impacts JSON-LD, Open Graph, and SEO meta tags.
Three Levels of YAML Frontmatter Fields
Required → Recommended → Advanced
Required: 4 Fields
title, description, slug, date — the build will fail without them. Filled out in 30 seconds. They form the H1 heading, meta description, page URL, and publication date in sitemap and JSON-LD.
Recommended: 6 Fields
author, category, schema_type, tags, layout, cover_image. They affect JSON-LD Person, E-E-A-T, categorization, and OG image. +40-60% AI citation rate.
Advanced: 5+ Fields
raw_html, geo, certifications, sources, updated, lastReviewed. For GEO optimization, E-E-A-T signals, and Featured Snippets. Used in expert deep-dive articles.
How Frontmatter Becomes JSON-LD and SEO
gray-matter parses YAML
Block between --- in .md file → object with fields. 0.1 milliseconds to parse.
useSeo.ts → JSON-LD + OG
title → og:title + headline. author → JSON-LD Person. schema_type → @type.
npm run build → everything in HTML
JSON-LD, OG, meta, sitemap — generated at build time. Without plugins.
6 Key Fields and Their Impact on SEO/GEO
How each field works in JSON-LD and meta tags
→ og:title + headline
Page H1 + JSON-LD headline + og:title. Maximum 70 characters. The primary SEO factor for content.
→ meta + og:description
Meta description + JSON-LD description. 120-160 characters. Snippet in Google and AI search engines.
→ Page URL
Clean URL: /blog/my-slug/. Only Latin characters, numbers, hyphens. Validated at build — will fail with spaces.
→ JSON-LD Person
Author name. Automatically generates JSON-LD Person with link to Organization. E-E-A-T signal for AI.
→ JSON-LD @type
Article, HowTo, FAQPage, Product. Defines Schema.org type for JSON-LD. Core GEO optimization field.
→ sitemap lastmod
Publication date in YYYY-MM-DD format. Shown in sitemap datePublished in JSON-LD. Affects content freshness.
Three Categories of Fields: From Basic to GEO
How fields are grouped by purpose
SEO Fields
title, description, slug, date, updated — form meta tags, canonical URL, sitemap lastmod, and og:title/description. This is the foundation: the page cannot be indexed properly without them.
- Required minimum: title, description, slug, date
- updated — last modified date (sitemap)
JSON-LD Fields
author, schema_type, tags, category, certifications, sources — directly feed into Schema.org JSON-LD. AI reads who the author is, what the category is, and how trustworthy the source is.
- author → Person, schema_type → @type
- certifications + sources → citedBy
GEO Fields
raw_html, geo, cover_image, layout — for GEO optimization and visual appearance. raw_html: true allows inline HTML styles. geo adds geo-coordinates to JSON-LD.
- raw_html: true — use article templates
- geo — coordinates for local SEO
Evolution of YAML Frontmatter
5 stages of metadata development for static sites
2008 — Jekyll and the Emergence of Frontmatter
Jekyll introduces YAML Frontmatter for static sites: a block between --- with metadata. 3 fields: title, layout, date. The standard for all SSGs.
20082014 — gray-matter and Node.js
The gray-matter library for Node.js: parsing YAML/JSON/TOML Frontmatter. Used in Metalsmith, Gatsby, Hexo. 10+ fields, arbitrary metadata keys supported.
20142018 — Schema.org and JSON-LD
Google recommends JSON-LD structured data. Frontmatter starts including schema_type, author, certifications for automated @type generation.
20182024 — GEO and E-E-A-T Signals
AI search engines (SearchGPT, Perplexity) require E-E-A-T signals. Frontmatter adds geo, certifications, sources — fields for AI citation.
20242026 — GitHub CMS: 15 Fields + Auto JSON-LD
15 YAML fields. 10+ Schema.org types automatically generated. @block directives for AI. Pipeline validation at build. Content metadata standard for the AI era.
2026How author and schema_type Become JSON-LD Person
The author field in YAML Frontmatter automatically generates JSON-LD Person linked to Organization. schema_type determines @type: Article, HowTo, FAQPage. AI sees: who wrote it, what structure, whether it can be trusted.
from author
from schema_type
4 Steps to Fill Out Frontmatter
From empty file to AI-optimized page
Create .md file
content/blog/article.md or content/sections/name.md. Add --- at the top.
30 secFill required fields
title, description, slug, date. 4 fields — the foundation. Build will fail without them.
30 secAdd recommended
author, category, schema_type, tags. +40-60% AI citation rate. JSON-LD Person auto-generated.
30 secnpm run build
Build converts YAML to JSON-LD + OG. Validation catches errors. 10 tests pass — page ready.
60 secTestimonials: YAML Frontmatter in Real Projects
Results from using GitHub CMS Frontmatter
Alexey K.
CEO B2B Platform
"57 pages — filled Frontmatter in an hour. JSON-LD generates automatically. Before, every WordPress plugin required manual configuration. author → Person, schema_type → Article — all from one YAML block."
Marina S.
Tech Director EdTech
"Frontmatter is brilliantly simple. Fill 15 fields in 2 minutes and forget. useSeo.ts generates OG tags, JSON-LD, sitemap automatically. Build validation catches errors: slug with spaces, date not in format. WordPress can only dream of this."
Dmitry V.
DevOps Engineer
"Content validation at build — 10 tests. If slug has spaces, the build fails. If date is not YYYY-MM-DD — the build fails. WordPress publishes with errors. Here — everything is checked before production."
FAQ on YAML Frontmatter
Common questions about filling out metadata
Frontmatter → JSON-LD: Complete Chain
YAML → gray-matter → object
--- block is parsed into a JavaScript object. All fields available as parsed.data.
Object → useSeo.ts → JSON-LD
author → Person, schema_type → @type, tags → about. 10+ Schema.org types.
JSON-LD → <script> in HTML
Embedded in the head of every page. AI crawlers parse directly in 0.3 seconds.
6 Advantages of YAML Frontmatter Over WordPress
Why Markdown metadata beats plugins
15 fields → auto-JSON-LD
Without plugins. WordPress: Yoast + Schema Pro ($168-728/year).
Build validation — 10 tests
WordPress: errors are discovered in production.
Content in Git — full history
WordPress: revisions in MySQL (incomplete).
$0 vs $168-728/year
GitHub CMS: everything built-in. WordPress: 3-5 paid plugins.
AI-readable — 0.3s parsing
WordPress: AI chokes on junk HTML.
2 minutes — 15 fields
WordPress: Yoast + Schema Pro = hours of setup.
Fill Out Frontmatter in 2 Minutes — Get JSON-LD Out of the Box
Clone GitHub CMS — 15 YAML Frontmatter fields generate everything: SEO meta tags, 10+ JSON-LD types, Open Graph, sitemap. Without plugins, without subscriptions.
Free · MIT License · 15 fields · 2 minutes · → JSON-LD
Article from Section 3: Content / Markdown. Created using prompt template article-2.txt (HOME-4 style). YAML Frontmatter — Complete GitHub CMS Field Reference Guide.