SEO, sitemap, and llms.txt

Public crawl and LLM discovery surfaces are generated from the same site manifest and content sources.

Page metadata

src/lib/seo.ts builds shared head payloads:

• title / description
• canonical URL
• Open Graph + Twitter cards
• optional article:* fields for blog posts
• noindex for private or auth surfaces when needed

Public pages should call buildPageHead({ title, description, path, ... }) from their route head() function.

Sitemap and robots

• /sitemap.xml lists static public routes, docs pages, and published blog posts when the blog service is reachable
• /robots.txt allows public paths, blocks auth/dashboard/API, and points to the sitemap + llms index

Implementation: src/lib/server/public-routes.ts plus route handlers under src/routes/.

llms.txt

• /llms.txt is a short index of core pages, docs, blog, and legal routes
• /llms-full.txt expands the same set with full docs/blog body content when available

These files are designed for LLM crawlers and internal agent tooling. Keep them public and free of secrets.

Content ownership

• Docs catalog: src/lib/content/docs.ts (mid-term can move to markdown files)
• Blog posts: BLOG_SERVICE API (cover, author, date, description, markdown)
• Legal pages: locale legal copy modules