Search Tips

1 min read (154 words)

The search box in the site header is client-side: Kiln builds a JSON index of every published page at dist/search-index.json, ships a tiny script alongside it, and the browser does the filtering. No server, no service, no tracking.

How the index is built

Each entry carries a title, the page URL, a plain-text excerpt, and the post's tags. Because the index is generated after the publish filter, drafts and future posts are absent from it — you cannot find what was never shipped.

Getting good results

What it will not do

The demo script has no fuzzy matching and no stemming — it favors predictability over cleverness. For a site of this size, that is plenty.

guides