Search and AI visibility audit

VNZT search and AI citation readiness report

Audit of technical SEO, AI crawler access, schema, Knowledge Graph, llms.txt, E-E-A-T, content and readiness for AI platforms.

Executive summary

VNZT now looks less like a conventional site and more like a knowledge framework: human-readable pages, AI Layer, Knowledge Graph, API resources, documentation, versions, project status and llms files work together.

After fixes, the site is accessible to search and AI crawlers, has a clean sitemap without JSON/Markdown, JSON-LD coverage for HTML pages and a synchronized entity graph.

Main conclusion: the technical and AI-readable foundation is strong. The next growth area is external entity evidence: independent mentions, publications, author/project profiles and citeable research artifacts.

Scores by area

Technical SEO/GEO92/100StrongSitemap URLs return 200, HTTPS, canonical URLs, headers, robots and AI crawler access work. Remaining issue: HTTP to HTTPS is still a 302 at hosting level.
Schema / Knowledge Graph90/100StrongHTML pages contain JSON-LD, stable @id values are used, graph/entities/relations are synchronized.
AI Visibility / llms88/100Strongllms.txt and llms-full.txt describe the platform, model, AI resources, interpretation rules and research status.
Content / E-E-A-T74/100Good but needs strengtheningThe research frame is clear, but external references, protocols and validation artifacts need further development.
Brand / Entity Authority58/100Needs workFirst-party signals are strong, but independent external evidence is still limited.
Platform readiness82/100GoodGoogle, Gemini, Perplexity, Bing, ChatGPT and Claude can technically access key pages; citation authority is the next step.

Fixed after audit

GPTBot, ClaudeBot and bingbot received 403FixedThey now receive 200 on key HTML and JSON pages.
Security headers were missingFixedHSTS, CSP, X-Content-Type-Options, X-Frame-Options, Referrer-Policy and Permissions-Policy were added.
JSON-LD coverage was sparseFixedAll sitemap HTML pages include JSON-LD.
Knowledge Graph projections were inconsistentFixedgraph.json, entities.json and relations.json are synchronized.

Remaining gaps

The main gap is external authority. AI systems can read VNZT, but they still need corroborating signals: public profiles, repositories, OSF/Zenodo artifacts, preprints, citations or independent discussion.

The second gap is empirical validation. H1, H2 and H3 are promising because they can fail, but they still require preregistration, sample sizes, controls, analysis plans and published results.