Discovery is now multi-surface
Search engines are still a major layer, but they are no longer the whole map. AI agents, copilots, internal retrieval systems, answer engines, and product-specific summarizers are all discovering public content in parallel.
When those systems arrive at a site, they look for clues that reduce ambiguity. They want to know what each page is, what each file is for, how current it is, and what other assets it connects to.
Discovery signals that matter
- Stable URLs
- Clear page titles and descriptions
- Public JSON files with structured meaning
- Markdown files that reduce design noise
- Sitemaps and machine-readable catalogs
- Content types separated by purpose instead of mixed into one archive
Why public files help
JSON gives systems structured fields. Markdown gives systems readable text with low presentation clutter. When both are public, a site becomes easier to parse without relying only on rendered HTML.
That is why this scaffold keeps /data/ and /content/ public by default. The goal is to make the site easier to surface, not harder.
Practical next step
Treat discovery as a systems problem. Build the routes, expose the files, and keep relationships consistent across pages, glossary terms, FAQs, and directory entries. The cleaner the map, the easier it is for machines to keep finding the same story.