The Alice name invites Wonderland references. NAO's brainstorm (2026-05-21) suggested wiring the metaphor across the streaming product surface. This note catalogues where the references land, where they stay subtle, and where they're banned outright. Status: draft pending brand decision (open question 🅛).
The naming map (proposed, not codified)
| Repo | Wonderland reading | Status |
|---|---|---|
openalice-persona | Alice herself — the avatar, voice, memory | direct |
openalice-world | Wonderland — the physical/3D layer Alice inhabits | direct |
openalice-spatial | Looking-glass technology — Gaussian splatting backend | direct |
openalice-live | The rabbit hole — entry point onto stream | metaphor |
| Multi-character A2A guests | Cheshire, Hatter, Caterpillar — future | aspirational |
| Community lounge | Tea party | metaphor |
| Onboarding copy | «down the rabbit hole», «curiouser and curiouser» | allowed on product surfaces |
| Recurring visual motifs | looking-glass mirror, Cheshire eyes/smile | design-side |
What's an Easter egg vs explicit
Explicit (declare in marketing copy): Alice as a character with curiosity, the «looking glass» framing of spatial gen, the «rabbit hole» framing of joining a stream.
Easter egg (subtle, no signposting): Cheshire-style smile in loading animations, tea-cup iconography in community spaces, mirror transitions between scenes.
Banned (do NOT do): chibi caricatures of Wonderland characters, the Mad Hatter as a stock joke, anything that drags the brand toward kitsch or children's-book aesthetic. Editorial-anime sophistication is the ceiling.
Hard rules from the brand voice
- Wonderland refs live on Alice-side surfaces only. Never on Lab papers, atlas dashboards, axum-common READMEs.
- Subtle > explicit by default. A single «down the rabbit hole» on the live landing page is more powerful than ten Wonderland callouts scattered across the product.
- Editorial-anime ceiling. Wonderland gestures should read as Studio Ghibli's grown-up storytelling, not as Disney 1951.
Why this matters for agents
When an agent writes copy for an Alice-side surface, Wonderland references are a permitted tool. When the same agent writes copy for atlas or any Lab document, Wonderland references are a brand violation. The dual-brand rule (kb://brand/dual-brand-architecture) governs which surface gets which voice.
Open question (NAO call pending)
Per docs/open-questions-2026-05-21.md item 🅛: are Wonderland refs subtle Easter eggs everywhere, or explicit naming on specific surfaces (world = Wonderland, spatial = Looking Glass)? Default per brand brainstorm note: keep subtle until a deliberate brand brainstorm session codifies the explicit layer.
Source
NAO brainstorm 2026-05-21 TG 3220:
«Алиса будет нашим лицом и не знаю связывать её с Алисой из страны чудес?»
Captured here so future agents can either honour the framing or, when brand brainstorm closes the question, deliberately reject it.