Search
Search the corpus Full-text search across specifications, insights, compliance pages, the roadmap, and the manifesto. Pagefind indexes the published site at build time; results link to the live page.
Search requires JavaScript. With JavaScript disabled, the Vault index , references graph , and llms.txt all provide navigational surfaces over the same corpus.
What's indexed The full body text of every published specification (15 at M7 ship) and every per-spec changelog route; every insight article; every compliance page; the roadmap, the framework architecture page, the manifesto, and the about page. Pagefind indexes the rendered HTML, so cross-reference resolution, frontmatter, and headings are all searchable.
Results rank by relevance, with title matches weighted higher than body matches. Pagefind ignores HTML chrome, code blocks where the page disables indexing via data-pagefind-ignore, and pages explicitly marked excluded (the search route itself is excluded).
Search indexes regenerate on every build. If a recently updated spec doesn't surface here, the build hasn't shipped yet.