In plain English
This page is part of the technical reference. It keeps the expert detail but starts with a plain-language summary for first-time readers.
- Why this matters: AI risk can come from the whole arrangement, not one obvious model.
- What to look for: data, memory, routes, adapters, tools, evaluators, updates, and rollback paths.
- Technical version below: the expert terminology remains available and is linked through the glossary.
Crawlable Metadata
The package renders essential page content server-side through PHP, includes unique metadata in app/content-index.json, emits canonical URLs, and generates a sitemap from indexed routes.
Direct answer
Cognivirus.com improves discoverability through ordinary technical publishing: unique titles, useful descriptions, stable extensionless URLs, canonical links, internal links, sitemap coverage, robots disclosure, and readable server-rendered HTML. It does not use hidden text, cloaking, prompt-injection instructions, or keyword-stuffed variants.
Current crawl controls
| Control | Implementation |
|---|---|
| Page title | Generated from content metadata and site name. |
| Meta description | Stored in the content index and visible in page source. |
| Canonical URL | Generated from the normalized route. |
| Sitemap | Regenerated from indexed routes and static discovery files. |
| Robots | Permits crawl of public pages and points to the sitemap. |
| Internal links | Validated by local tests against indexed routes. |
| Search index | Local JSON search index mirrors public route metadata. |
Repair made in v1.10.0
The previous package exposed too much site-publication guidance as public pages. This release moves the work into the site’s actual content, metadata, manifests, validation, and documentation. The public explanation is now about Cognivirus.com’s own publication integrity, not a copied guide.