Not an app. Not a browser. A world-rendered intent environment for accountable AI work — where you can explore what the AI is doing, see the gates it crossed, and verify the proof yourself.
Q navigates the work · Immaculate gates the boundaries · LaaS remembers every state change · the Verifier replays the proof · Arobi Network anchors chain-of-custody. Q is the first model in the HORIZON class, built by Arobi Technology Alliance; founder Gaetano Comparcola.
Fly the accountable-autonomy spine first-person: Q lights each governance gate, every passed gate crystallizes into a sealed evidence trail, and you stop at the Immaculate gate to Approve or Deny the action yourself — Deny breaks the chain and halts. Semantic zoom, click-to-inspect records (real LaaS event fields), reduced-motion aware. Self-contained, verifiable, honest fixture data.
Launch the immersive world →Watch a request flow the accountable-autonomy spine as a living river of work — Q lights each governance gate and leaves a sealed evidence trail. Click any gate to inspect its record.
The web as a navigable landscape: Q tours the diagnosis districts, pins evidence with source support, then Immaculate gates the real-world action before it runs. The PRD's §6.1 research journey, made interactive.
Paste a signed /v1/xena/capabilities or /v1/xena/assets response. Your browser recomputes the digest and checks the Ed25519 anchor with the public key alone — trusting no Arobi server.
The lighter-weight 2.5D <canvas> map of the same fluid-intent fixture — a replayable path through typed nodes and governed gates. The earlier renderer, kept as the minimal variant.
The public marketing surface — now carrying the HORIZON story: "Use Q — Arobi's own reasoning model, the first in the HORIZON class." The private Q gateway customers connect to.
↓ Download the XENA OS bundle · static · self-contained · v0.1.0 · live