Search

Search the site

Use Search in the header, then type words from a page title, section heading, or article. Results link directly to the relevant page or heading. Press Escape or choose Close to return to the page.

Search runs in your browser using search-index.json, generated with the site. No search account, API key, or separate server is needed. The index loads when you open search, so normal page reading does not fetch it.

What matches

Matching is case-insensitive and supports partial words. Every query word must occur somewhere in an entry’s page title, heading, or section text. Title matches rank above heading matches, which rank above body-only matches.

The index is rebuilt with the content. Production builds exclude drafts; development preview includes them. Pages omitted from explicit sidebar navigation remain searchable.

Choose guide or API scope to narrow results. Matching words are highlighted; recent searches remain in local browser storage and can be cleared in the dialog. Use search: false in frontmatter to exclude a published page from the index. noindex: true also excludes it from the sitemap, Markdown exports and AI indexes. Set search.enabled: false in configuration to omit the search control and index.

Read Markdown and AI exports for Copy Page and downloadable reading formats.

Limits and failures

Search requires JavaScript and browser support for the modal dialog. Without those, the pages and navigation remain readable, but Search is not shown.

This is a small-site search, not typo correction, stemming, or a hosted indexing service. There is no configuration for replacing the search provider.

If the index cannot load, check that search-index.json was deployed alongside the HTML and that the configured base path matches the published URL. Serve the output over HTTP(S), not by opening an HTML file directly from disk.