CMS Plugin · SEO

SEO — Search, Sitemap & Social

How the CMS makes every page search-ready: per-post meta fields, crawler prerendering, an automatic sitemap.xml and robots.txt, and Open Graph / JSON-LD structured data.

Search-ready out of the box

Every page and post the CMS publishes is SEO-complete the moment it goes live — no plugin, no theme tweak. Each one ships:

Full <head> metadata — title, description, canonical, robots, Open Graph, Twitter cards and JSON-LD.
A static prerender for crawlers — search bots get fully-rendered HTML, humans get the live SPA.
Automatic sitemap.xml and robots.txt — generated from the live content, tunable from the admin.

This page is itself served exactly this way: a static prerender for Googlebot, and the entry you can see right now in /sitemap.xml.

Per-page SEO fields

Each cms_post carries its own SEO metadata, edited in the page/post editor's SEO panel with a live search-result (SERP) preview and character-budget hints (title ≤ 60, description ≤ 160):

The SEO panel in the page editor — a live SERP preview over every editable field, with title/description length hints.

Field	Purpose
`meta_title`	The `<title>` and SERP headline (falls back to the post title).
`meta_description`	The search-snippet description.
`meta_keywords`	Legacy keywords tag (emitted; ignored by modern engines).
`og_title` / `og_description` / `og_image_url`	Open Graph fields for social share previews (title/description fall back to the meta pair).
`canonical_url`	Canonical URL; defaults to `{public_base_url}/{slug}`.
`robots`	The robots meta directive; default `index,follow`.
`schema_json`	Custom JSON-LD that overrides the auto-generated schema.
`seo_excluded`	“Exclude from search engines” — flips the page to `noindex` and drops it from the sitemap.

How search engines see your pages — prerendering

The public storefront is a Vue SPA, which a naive crawler would see as an empty shell. The CMS solves this with a prerender layer: when a post is published (or its content changes), a static ${VBWD_VAR_DIR}/seo/<slug>.html file is written with the post's content and a fully-built <head>. The front nginx then splits traffic by user-agent:

Crawlers (Googlebot, bingbot, …) are served the static prerender — complete HTML, no JavaScript required.
Humans get the live SPA, which re-syncs the same meta tags on the client (keyed by data-seo="ssr") so navigation stays correct.

The head builder injects the title, description, robots, canonical, Open Graph, Twitter-card and JSON-LD tags from the fields above. Prerenders refresh automatically on publish; an admin can also rebuild them all at once — the Prerendered content tab's Generate button calls POST /api/v1/admin/cms/seo/regenerate:

Admin → CMS SEO → Prerendered content — regenerate or clean up the static crawler HTML for every published post.

sitemap.xml

A standards-compliant sitemap is generated live at /sitemap.xml — no build step, no stale file. It lists only published, search-visible posts; anything with seo_excluded, a noindex robots value, or an excluded category is left out. Each entry carries <loc> (the canonical URL), <lastmod> (the post's updated_at), changefreq and priority, plus hreflang alternates for translations. Above 50,000 URLs it automatically becomes a sitemap index over /sitemap-<n>.xml chunks.

<url>
  <loc>https://vbwd.cc/docs-core-cms/seo</loc>
  <lastmod>2026-06-28T18:25:20</lastmod>
  <changefreq>weekly</changefreq>
  <priority>0.5</priority>
</url>

Which posts appear is tunable from the admin without touching code — include or drop pages, exclude specific slugs, or scope the sitemap to posts carrying (or not carrying) given categories:

Admin → CMS SEO → Sitemap.xml — narrow the published set by pages, excluded slugs, and include/exclude categories.

robots.txt

/robots.txt is served dynamically too. The default keeps crawlers out of the app surfaces and points them at the sitemap:

User-agent: *
Disallow: /dashboard
Disallow: /api
Disallow: /admin

Sitemap: https://vbwd.cc/sitemap.xml

The body is fully overridable from the Robots.txt admin tab (an empty box means “use the default template”). A site-wide kill switch, seo.mode = off, forces Disallow: / regardless of the custom text — handy for staging instances that must never be indexed.

Admin → CMS SEO → Robots.txt — override the served robots.txt; seo.mode=off forces a full Disallow.

Structured data & social cards

Beyond the basic tags, the head builder emits:

Open Graph + Twitter cards — og:title/description/image and a summary_large_image Twitter card, so links unfurl with a title, blurb and image on social platforms and chat apps.
JSON-LD — a schema.org WebPage (or Article) block is generated automatically; supplying schema_json on the post overrides it with your own structured data.
hreflang — translated posts that share a translation group cross-link with rel="alternate" hreflang tags plus x-default.

In short

Write a page, fill the SEO panel (or accept the sensible defaults), publish. The platform then: prerenders it for crawlers, lists it in sitemap.xml, respects it in robots.txt, and emits complete Open Graph, Twitter and JSON-LD metadata — with per-post overrides whenever you need them. See the CMS plugin reference for the content model these SEO fields hang off.