MindHYVE.ai

The MindHYVE family in five views

Six sites. 247 pages. Five ways to look.

MindHYVE.ai (the corporate parent) and five product surfaces — ArthurAI™, ChironAI™, JustineAI™, TheoAI™, and the IPSC corpus. Same data, five visualizations. Pick the one that answers your question.

776 cross-site links connect the six surfaces. Data generated by crawling each site’s public sitemap.xml.

01Five views, five questions

Each view answers a different question about the same family.

The data is the same: 247 pages, 776 cross-site links, six surfaces. The shape of the visualization changes what the data actually tells you.

02How this works

One data layer, five renderings. A small Node.js crawler (scripts/crawl-family.ts) fetches each family site’s sitemap.xml on every refresh, visits each canonical URL once to extract its <title> tag and outbound family links, and emits a single canonical JSON. All five visualizations read the same JSON and render it differently.

Static-export safe. No client-side network calls. Layouts that need d3 simulations (Galaxy, Chord) run on mount in the visitor’s browser; layouts that are deterministic (Constellation, Atlas, Tree) run once at React render time on the committed data. Either way, the JSON is the source of truth.

Last refresh: May 25, 2026, 02:55 PM PDT.