Skip to main content

BYO-screen family display (kiosk mode) — design sketch

Date: 2026-06-24 · Status: Sketch / direction (not scheduled). Sibling to T7 (docs/superpowers/specs/2026-06-24-t7-household-restructure-design.md) and the North Star tv_dashboard entitlement. Strategic rationale: docs/market-analysis-2026-06.md.

What it is​

A landscape, always-on, glanceable "family display" MODE of the Rewhaven app — the tv_dashboard entitlement — that runs on a screen the family already owns. It is the software answer to Skylight's wall display without Rewhaven manufacturing hardware.

Positioning — "we meet you where you are." The marketing message is the inversion of "buy our screen": Rewhaven runs on the devices already in the home — phone · tablet · laptop · desktop · Fire TV · Google TV · Apple TV. No new device; put it on the TV you already have. (See market analysis §12.)

Why (and why it matters more for the ICP)​

Skylight's one real edge is the always-on, glanceable ambient surface. For neurodivergent kids, an always-visible visual schedule is not just calendar parity — it's a genuine executive-function affordance (externalizing "what's now / what's next" reduces the working-memory load that ADHD/autism tax). So the family display is part of the affirming product, not a bolt-on. It also creates the shared, neutral, phone-free surface kids without phones can use.

What it shows (glance hierarchy, big + legible across a room)​

  1. Per-member "Now / Next" lane — each kid as a column/row: their current chore (Now) and the next one (Next), first/then style. Reuses the Today wiring (T6): assignment filter + per-member resolved steps.
  2. Token + goal progress — each kid's give/save/spend glance (walletOf) + their top individual goal progress bar (T7). A Family goal progress bar across the top (T7 scope: family).
  3. Approvals waiting — a subtle "N waiting for a grown-up" chip; a parent glances and taps to approve (or approves on their phone).
  4. Idle screensaver — rotating family photos when no one's interacting (cheap Skylight parity; reuses nothing domain-side).
  5. Later: the events-calendar overlay (the deferred calendar expansion) — chores
    • goal deadlines + events on one grid.

Interaction model​

  • Primarily glanceable / read-only ambient. Most of the value is passive.
  • Light touch: a kid taps their chore → marks done → routes through the normal ChoreSubmission → human-in-the-loop approval (no punitive deductions; a parent approves here or on their phone). Big touch targets, high contrast, large type.
  • Kiosk lock: stays on the display screen, no nav-away, screen-keep-awake.
  • Shared-surface auth: the wall device shows household-scoped data (no per-user login needed to view); sensitive actions (approve, edit) gate behind a parent PIN or happen on the parent's phone. Ties to roles (admin/helper approve; MemberKind.isParental gates).

Technical approach (low domain cost — mostly a new layout)​

  • Flutter is already multi-platform → add a landscape "display" layout variant
    • a dedicated kiosk route, gated by the tv_dashboard entitlement. It is mostly presentation over existing data: Today (T6), wallets (walletOf), goals (T7), approvals — little/no new domain.
  • BYO-screen targets (honest reach per platform):
    • Phone / tablet — native (Android, iOS); a cheap Fire/Android tablet on a $15 stand or an old iPad in kiosk/lock-task mode is the canonical wall device.
    • Laptop / desktop — Flutter web (any browser, full-screen) + native desktop. Easy.
    • Fire TV / Google TV / Android TV — native Android build (D-pad/focus handling needed), or cast the web build via Chromecast/Cast.
    • Apple TV — no native Flutter tvOS embedder today; reach it via AirPlay-mirror from an iPhone/iPad/Mac (or a later dedicated tvOS effort). Market it as supported; deliver via AirPlay/cast first.
    • The web build is the universal fallback — anything with a browser or a castable surface. Offline-first: the display is just another SDK client of the local (or synced) store.
  • Keep-awake + auto-dim on idle; reconnect-resilient (Wi-Fi blips shouldn't blank it — a softer failure than Skylight's connectivity fragility).
  • Setup flow: a "Set up your family display" guide + affiliate links to a specific recommended cheap tablet + stand (no inventory).

Relationship to T7 / sequencing​

The display surfaces T7's outputs (member goals, family goals) + T6's Today + the wallet glance. So it is best built after T7 (so goals/profiles exist to show), but a thin first cut (per-member Today + token/goal glance, read-only) is feasible earlier since Today + wallets already exist.

Suggested phasing (when scheduled):

  1. P1 — read-only glance: per-member Today lane + token/goal glance, landscape kiosk on web/tablet. Entitlement-gated.
  2. P2 — light touch: tap-to-done (→ submission) + tap-to-approve (PIN); keep-awake
    • kiosk lock.
  3. P3 — ambient polish: idle photo screensaver, auto-dim, family-goal banner.
  4. P4 — calendar overlay (with the deferred Event domain).

Out of scope (for the sketch)​

  • First-party hardware (avoid; see market analysis §12).
  • The events calendar itself (separate deferred slice).
  • Multi-household / cast-account complexity.

Open questions (decide if/when scheduled)​

  • Default landscape breakpoint + min legible distance (TV across a room vs counter tablet) — affects type scale + density.
  • Parent-PIN vs phone-only for approvals on the shared surface.
  • Whether the thin P1 cut ships before T7 (Today-only) or waits for goals.