SEO and Feeds

A default build already emits meta tags, structured data, a sitemap, robots.txt, llms.txt, feeds, and social card images. The seo, sitemap, robots, rss, atom, llms_txt, and social_cards plugins are all enabled by default, so the work left is setting the site URL and per-page descriptions and images. Turn any of them off in Plugins and checks.

Set site.url in sarde.yaml to the address the site is published at. Canonical links, sitemap entries, feed links, and social card URLs are built from it:

site:
  url: "https://example.com"

Meta tags

The seo plugin adds Open Graph and Twitter Card meta tags to every page:

• Open Graph: og:title, og:description, og:url, og:type, og:site_name, og:locale, and og:image with its alt text. Pages in a collection get og:type set to article plus article:published_time and article:modified_time. On multi-language sites, og:locale:alternate lists the translations.
• Twitter Card: twitter:card (default summary_large_image), twitter:title, twitter:description, twitter:image, and twitter:site when a handle is configured.

Values come from the page's frontmatter and the site configuration. Set the description and image per page:

---
title: Photosynthesis
description: "How plants convert light into chemical energy."
image: /images/photosynthesis.png
---

The image resolves in this order: the page's image, a generated social card, then the default_image option.

When description is not set, the plugin uses the page summary (the first prose paragraph, truncated; code fences and directive blocks are skipped). If the body has no prose at all, such as a homepage built only from directive blocks, it uses text taken from the rendered page. Set auto_description: false to turn the fallback off.

To control indexing for one page, set robots in its frontmatter:

---
title: Internal Notes
robots: "noindex,nofollow"
---

Pagination pages after the first (/blog/page/2/ and later) get noindex,follow automatically and are left out of the sitemap.

Site-wide SEO options

Set the Twitter handle, card type, and a fallback image in the seo plugin config:

plugins:
  config:
    seo:
      twitter_handle: "@getsarde"
      twitter_card: "summary_large_image"
      default_image: "/images/og-default.png"

See SEO for every option and for the full list of generated tags.

JSON-LD structured data

The seo plugin writes a <script type="application/ld+json"> block containing a @graph of these nodes:

The BreadcrumbList follows the visible breadcrumbs, including collection and section titles. Article reads its author from params.author, and Course reads its provider from params.provider. Set all three under params in frontmatter:

---
title: Photosynthesis
params:
  schema_type: Course
  provider: "Riverside High School"
  author: "Dr. Chen"
---

Canonical URLs

Every page gets a <link rel="canonical"> tag with its absolute URL. In versioned documentation, pages of older versions point their canonical to the matching page in the latest version, which consolidates search ranking on the current docs.

hreflang alternates

On multi-language sites, the default theme adds a <link rel="alternate" hreflang="..."> tag for every translation of a page, plus an x-default entry for the default language. Search engines use them to serve the right language version.

Social card images

The social_cards plugin generates a 1200x630 PNG for each page that has no image of its own, and uses it as og:image and twitter:image. A card shows the page title, description, site name, the collection name, and, for pages with an explicit date, the date, in the active theme colors. A logo mark and a watermark are optional.

Cards are written to og/ during sarde build. sarde dev does not write them. The plugin depends on seo, which supplies the title and description tags that accompany the image.

plugins:
  config:
    social_cards:
      skip_if_image: true
      format: "png"

When skip_if_image is true (the default), a page with an image in its frontmatter uses that image and gets no card. format also accepts jpeg, which adds the quality option. See Social Cards for colors, logo, background, fonts, and per-page overrides.

Sitemap

The sitemap plugin writes sitemap.xml at the site root. It lists every published page with a lastmod date and leaves out drafts and pagination pages. changefreq and priority are configurable, and exclude takes URL patterns to skip.

See Sitemap for the options.

RSS and Atom feeds

Blog collections generate feeds without configuration:

• RSS 2.0 at /<collection>/feed.xml
• Atom 1.0 at /<collection>/atom.xml

Each feed holds the 20 most recent posts by default. An entry carries the title, link, date, and a summary taken from the page description, falling back to the page summary. Feeds do not include the full post body.

Set feed: false on a collection to turn its feeds off, or feed: true on another collection to add feeds for it. See Feeds for the entry limit and the collections option.

robots.txt

The robots plugin writes robots.txt with an Allow: / rule. It adds a Sitemap: line pointing to sitemap.xml only when the sitemap plugin is enabled and site.url is set.

See Robots for the options.

llms.txt

The llms_txt plugin writes /llms.txt, a Markdown index of the site for AI tools. It lists each content page with its title and URL and leaves out section index pages and the homepage. Configure it under the top-level llms_txt key, not under plugins.config. See LLMs.txt for the options.