Content

svedocs reads .md, .mdx, and .svx files from the configured content folders. Each file becomes a page with a route, navigation entry, search records, and SEO data.

Route mapping

The default content roots are:

Text
1content/docs -> /docs2content/pages -> /

Examples:

File	Route
content/docs/index.md	/docs
content/docs/writing/content.md	/docs/writing/content
content/docs/configuration/index.md	/docs/configuration
content/pages/changelog.md	/changelog

Use index.md for a section landing page. Use nested directories when the sidebar should have groups.

Multilingual projects place the locale path below each content root. See Internationalization for the complete file-to-route mapping.

Frontmatter

Every important page should include at least title and description:

Markdown
1---2title: Content3description: Explain the content pipeline and authoring features.4order: 25---

Common fields:

Field	Type	Purpose
title	string	Page title, search title, OG title, and default sidebar title.
navTitle	string	Shorter sidebar label without changing the document title.
description	string	SEO description, search excerpt fallback, and page summary.
order	number	Sort weight in navigation and previous/next links.
hidden	boolean	Remove from generated navigation and public lists.
collapsed	boolean	Collapse a navigation group by default.
section	boolean	Mark a directory page as a section page.
icon	string	Icon hint for the default theme.
canonical	string	Override the generated canonical URL.
image	string	Open Graph image URL.
keywords	string[]	SEO and search metadata.
author	string	Article author; falls back to seo.defaultAuthor.
published, date, publishedTime	date	Publication time.
updated, updatedTime	date	Last meaningful content update.
type, ogType	string	Open Graph content type.

If the first Markdown heading matches title, svedocs removes the duplicate heading from rendered content while preserving the page title.

Markdown features

svedocs supports GitHub-flavored Markdown features such as tables, task lists, and autolinks. It also extracts headings for anchors and page table-of-contents data.

Markdown
1## Configure search23- [x] Create the runtime route.4- [x] Choose a provider.5- [ ] Add production credentials.

KaTeX is available for inline and block math when your content needs formulas.

Code blocks

Use code metadata to make examples easier to scan:

```ts title="svedocs.config.ts" {1} focus=3
import { defineConfig } from 'svedocs/config';

export default defineConfig({
  site: { name: 'svedocs' }
});
```

The default theme supports:

• Syntax highlighting.
• File titles with title="..." or filename="...".
• Line highlighting with {1,3-5}.
• Focus hints with focus=3.
• Optional line numbers and wrapping from theme config.
• Copy actions.

Diff blocks

Use standard diff fences for small changes:

Text
1- const docs = fragmented();2+ const docs = svedocs();

Use split diffs when the before/after relationship matters:

Diff+1-1
-@@ -1,3 +1,4 @@+-export * from './content.js';3+export * from './core/content.js';4+export * from './core/navigation.js';5 export * from './config.js';

Split diffs support horizontal scrolling on both panes, so long code stays inspectable on narrow screens.

Links and assets

Internal links are checked against generated routes and extracted heading anchors:

External links receive a small inline icon in the default theme so readers can distinguish off-site navigation:

Markdown
1Read the [SvelteKit docs](https://svelte.dev/docs/kit).

For example, the SvelteKit documentation link is marked as an external web link.

Use a standalone internal link with a card title to render a Fumadocs-style link card:

Markdown
1[SEO and OG](/docs/integrations/seo-og "card: Metadata, sitemap, robots, JSON-LD, and Open Graph routes.")

Local assets are checked when checks.assets is enabled:

Markdown
1![Dashboard screenshot](/images/dashboard.png)

Use absolute site paths for public assets and stable docs routes for internal links. This keeps links working in search results, Ask AI citations, and static builds.

Search records

svedocs creates two kinds of search records:

Record	Created from	Useful for
Page records	The page title, description, route, and content.	Broad queries that should land on the page.
Section records	Headings and nearby content.	Specific queries that should jump into a long guide.

Clear headings improve search quality more than keyword stuffing. Prefer headings that describe a user task or concept.

Authoring checklist

Before publishing a page:

• Does the page have a unique title and useful description?
• Does the route belong in docs or should it be a standalone page?
• Are headings specific enough to become search targets?
• Do code examples include file names when context matters?
• Do internal links point to real routes and anchors?
• Does the page need canonical, image, published, or updated metadata?