Eve library

OSHUN V1 Customer Shell Redesign Inventory

Guide

Revision 1

On this page

Status: Accepted v1.0
Owner: OSHUN Design + Product + Engineering
Last Updated: March 27, 2026
Implements: V1-CSH-061
Related:

  • docs/releases/v1/specs/todos.md
  • docs/releases/v1/reports/current-state-redesign-inventory.md
  • docs/releases/v1/specs/redesign-sequencing-plan.md
  • docs/releases/v1/design/customer-visual-thesis.md
  • docs/releases/v1/specs/surface-redesign-ranking.md
  • docs/releases/v1/specs/experience-scorecards.md

1. Purpose#

This document is the redesign inventory for the launch-critical customer shell surfaces called out in the V1 consumer-shell overhaul:

  • home
  • explore
  • activity
  • library
  • notifications
  • profile
  • onboarding
  • auth

The goal is not to restate that these routes exist. The goal is to capture how they currently read, where the visual and structural problems are, which jobs must survive redesign, and which platform gaps will constrain the next redesign tasks.

2. Inputs Reviewed#

The inventory below is grounded in the current repository implementation on March 27, 2026.

Primary code inputs reviewed:

  • web shell routes in apps/oshun/web/src/app
  • web customer shell primitives in apps/oshun/web/src/components
  • mobile tab routes in apps/oshun/mobile/app/(tabs)
  • mobile auth/onboarding surfaces in apps/oshun/mobile/src/auth
  • shared shell chrome in apps/oshun/web/src/components/ShellLayout.tsx
  • current-state inventory in docs/releases/v1/reports/current-state-redesign-inventory.md
  • existing route audits already covering home, explore, activity, library, notifications, profile, and onboarding behavior

Visual evidence reviewed:

  • apps/oshun/web/e2e/visual-regression-pages.spec.ts-snapshots/page-home-desktop-chromium-darwin.png
  • apps/oshun/web/e2e/visual-regression-pages.spec.ts-snapshots/page-profile-desktop-chromium-darwin.png
  • apps/oshun/web/e2e/visual-regression-pages.spec.ts-snapshots/page-notifications-drawer-desktop-chromium-darwin.png
  • apps/oshun/web/e2e/visual-regression-pages.spec.ts-snapshots/page-onboarding-desktop-chromium-darwin.png

3. Redesign Frame#

3.1 Visual Thesis#

OSHUN should feel like one calm, nocturnal, editorial shell with Tara-led breathing room at the center, not a stack of adjacent domain dashboards sharing the same dark paint.

3.2 Content Plan#

Every redesigned surface should answer the same sequence in a route-appropriate way:

  1. orient the user immediately
  2. present one dominant next move
  3. show supporting context without turning into a module pile
  4. provide a clear continuation or governance handoff

3.3 Interaction Thesis#

The shell should reduce cognitive load through:

  • one focal plane per route instead of many equivalent panels
  • shared route transitions and reveal rhythm across web and mobile
  • calmer, lower-friction entry into assistant, notifications, and cross-domain handoffs

4. Coverage and State Labels#

This inventory covers both web and mobile when a surface exists on both platforms.

State labels used here:

  • implemented: dedicated route or primary surface exists
  • embedded: capability exists, but only inside another surface
  • missing: launch-critical surface job exists, but no dedicated implementation exists on that platform

5. Surface Summary#

Surface Web state Mobile state Current posture Priority Primary redesign blocker
Home implemented implemented Broad continuity dashboard P0 First viewport and below-the-fold rhythm are still module-dense
Explore implemented implemented Discovery workspace P0 Search, curation, and route guidance compete at the same hierarchy
Activity implemented implemented Timeline plus inbox hybrid P1 Progress, celebration, and triage live in one visual layer
Library implemented implemented Saved-item preserve surface P1 Strong information density still resolves as stacked cards and helper panels
Notifications embedded + overlay implemented + supporting settings Mixed-context inbox P1 Triage logic is sound, but the presentation is drawer-heavy and filter-dense
Profile implemented implemented Governance and account control surface P1 Long accordion stack reads as settings exhaust rather than guided governance
Onboarding implemented embedded Setup wizard P0 Web has a clear route, mobile does not
Auth implemented as public welcome/auth entry embedded in mobile profile Public entry plus account recovery P0 Web and mobile auth posture do not match each other or the shell tone

6. Detailed Inventory#

6.1 Home#

Entrypoints:

  • web: apps/oshun/web/src/app/page.tsx, apps/oshun/web/src/components/home/*
  • mobile: apps/oshun/mobile/app/(tabs)/index.tsx
  • related overlay: apps/oshun/web/src/components/NotificationsCenterPanel.tsx

Current implementation:

  • web home layers a hero, continuation rails, mixed-domain recommendations, quick actions, companion sections, direct-launch domain cards, KPIs, routine blocks, activity, wearable summaries, and notification preview in one route
  • mobile home similarly combines metrics, continuity, recommendations, domain cards, quick actions, library saves, and production-tool entry points

What is already strong:

  • Tara-led centering intent is present in copy and route logic
  • cross-domain continuity is meaningfully real, not placeholder IA
  • home is clearly the highest-value orientation surface in the shell

Redesign liabilities:

  • the route still resolves as a tall series of bordered modules rather than one composed journey
  • too many blocks use equivalent card weight, so the eye never gets a dominant focus plane
  • mobile home repeats the same breadth problem in a smaller viewport, which raises thumb-travel and scanning cost

What redesign must preserve:

  • Tara-first centering
  • domain handoff readiness
  • mixed-domain continuation
  • quick access to the five core shell journeys

6.2 Explore#

Entrypoints:

  • web: apps/oshun/web/src/app/explore/page.tsx, apps/oshun/web/src/components/explore/ExploreDashboard.tsx
  • mobile: apps/oshun/mobile/app/(tabs)/explore.tsx

Current implementation:

  • web explore is a dense discovery dashboard with hero copy, search, filters, concept entry points, curated rails, collections, saved-state hooks, and operator-access leakage in the route body
  • mobile explore is a large single-file discovery surface with quick actions, search state, collection cards, domain switching, library preview, and thumb dock affordances

What is already strong:

  • real cross-domain discovery breadth exists
  • search and saved-state integration are already present
  • concept, collection, and spotlight primitives give redesign material to work with

Redesign liabilities:

  • explore still reads like a feature inventory rather than a browse posture
  • search, filter state, editorial discovery, and operational wayfinding all compete in the same visual layer
  • the web route explicitly includes operator-access messaging, which is a strong sign that shell boundaries are not visually resolved yet

What redesign must preserve:

  • blended discovery across all five domains
  • save-to-library behavior
  • domain-aware filtering
  • clear handoff from inspiration to action

6.3 Activity#

Entrypoints:

  • web: apps/oshun/web/src/app/activity/page.tsx, apps/oshun/web/src/components/activity/ActivityDashboard.tsx
  • mobile: apps/oshun/mobile/app/(tabs)/activity.tsx

Current implementation:

  • web activity is framed as a hero-led dashboard with timeline, achievements, continuity threads, streak, digest, and filter systems
  • mobile activity acts as both activity route and notification inbox, with optimistic mutations, offline queue support, filters, and grouped items

What is already strong:

  • continuity and re-entry are treated as product primitives
  • web and mobile both already understand activity as cross-domain, not per-domain
  • activity is connected to offline and network degradation, which matters for real-world reliability

Redesign liabilities:

  • web activity still uses glass panels and animated controls that feel like a component showcase
  • mobile activity conflates timeline reflection with inbox triage
  • celebration, unread urgency, and historical review are not clearly separated

What redesign must preserve:

  • cross-domain progress memory
  • lightweight re-entry into the last meaningful action
  • unread and priority signaling
  • offline-safe interaction behavior

6.4 Library#

Entrypoints:

  • web: apps/oshun/web/src/app/library/page.tsx, apps/oshun/web/src/components/library/LibraryDashboard.tsx
  • mobile: apps/oshun/mobile/app/(tabs)/library.tsx, apps/oshun/mobile/src/library/UnifiedLibraryView.tsx

Current implementation:

  • web library combines typed saves, collection groupings, grounding signals, research-practice rail, assistant hooks, and route handoff panels
  • mobile library presents a hero, stat strip, shell-relationship explainer, assistant CTA, and the unified saved-item view

What is already strong:

  • library already has a clear product job: preserve and relaunch
  • typed-save and provenance posture are much stronger than a generic bookmark list
  • research-to-practice continuity is explicit

Redesign liabilities:

  • web library is still expressed as layered cards, chips, and guidance panels
  • mobile library explains the shell relationship well, but the route still carries too much explanatory chrome before the actual preserve surface
  • preserve, triage, and relaunch are present, but not yet composed into one obvious narrative

What redesign must preserve:

  • typed saves and provenance
  • collection-level relaunch paths
  • research-to-practice continuity
  • assistant-assisted triage

6.5 Notifications#

Entrypoints:

  • web: apps/oshun/web/src/components/NotificationsCenterPanel.tsx
  • mobile: apps/oshun/mobile/src/notifications/NotificationCenter.tsx, apps/oshun/mobile/src/notifications/NotificationSettings.tsx
  • related surfaces: mobile activity inbox, web profile notification preferences

Current implementation:

  • web notifications live as a full or preview panel rather than a dedicated route, with mixed practice, support, and operator signals plus multiple filter dimensions
  • mobile has supporting notification center/settings components, while primary inbox behavior is absorbed into the activity tab

What is already strong:

  • mixed-context notification modeling is deliberate instead of naive
  • the product already distinguishes practice, support, and operator lanes
  • preview versus full variants show good intent about home not becoming the inbox

Redesign liabilities:

  • current presentation is filter-heavy before it is calm
  • the drawer posture makes the surface feel appended to home instead of native to the shell system
  • web, mobile, and profile notification governance are structurally related but visually fragmented

What redesign must preserve:

  • mixed-context triage
  • explicit support/operator handoffs
  • notification preview on home without home becoming a queue
  • per-domain and modality preferences

6.6 Profile#

Entrypoints:

  • web: apps/oshun/web/src/app/profile/page.tsx, apps/oshun/web/src/components/ProfileSettingsPanel.tsx, apps/oshun/web/src/components/profile/*
  • mobile: apps/oshun/mobile/app/(tabs)/profile.tsx

Current implementation:

  • web profile combines account summary, plan and billing, privacy, assistant controls, notification preferences, data rights, operator access, and session management inside one deep settings panel
  • mobile profile is similarly broad and additionally embeds mobile auth and mobile onboarding cards when those flows are needed

What is already strong:

  • governance depth is real and unusually complete for v1
  • session, billing, privacy, and export controls already exist across platforms
  • operator handoff is explicit instead of hidden

Redesign liabilities:

  • web profile reads as a long accordion and card stack rather than a guided governance surface
  • mobile profile is overloaded because auth, onboarding, preferences, security, and support all land in the same tab
  • settings density is high enough that “important” and “rarely needed” controls are visually flattened together

What redesign must preserve:

  • session/device management
  • billing and entitlements
  • privacy/data rights
  • explicit operator boundary

6.7 Onboarding#

Entrypoints:

  • web: apps/oshun/web/src/app/onboarding/page.tsx, apps/oshun/web/src/components/onboarding/OnboardingWizard.tsx
  • mobile: apps/oshun/mobile/src/auth/MobileOnboardingCard.tsx

Current implementation:

  • web onboarding is a true route-level wizard with named steps, domain choices, goals, interests, routine timing, notifications, accessibility, and readiness
  • mobile onboarding exists only as a card embedded inside profile-auth state

What is already strong:

  • the web onboarding first screen is one of the cleanest single-focus surfaces in the product today
  • onboarding already links to preference sync and route re-entry
  • the domain/goal model is thoughtful and rooted in actual shell jobs

Redesign liabilities:

  • web onboarding feels visually cleaner than the authenticated shell, which makes the product feel like two brands
  • mobile has no dedicated onboarding route, which weakens cross-device continuity and launch clarity
  • setup depth is strong, but there is still room to reduce the number of decision points before first value

What redesign must preserve:

  • domain selection
  • goal and interest framing
  • notification/accessibility setup
  • resume and re-entry handling

6.8 Auth#

Entrypoints:

  • web: apps/oshun/web/src/app/welcome/page.tsx, apps/oshun/web/src/components/welcome/WelcomePageView.tsx, apps/oshun/web/src/components/welcome/WelcomeAuthPanel.tsx
  • mobile: apps/oshun/mobile/src/auth/MobileAuthCard.tsx

Current implementation:

  • web auth is embedded into the public welcome experience and sits beside domain-marketing, feature proof, and distribution CTAs
  • mobile auth is not a route; it is a card that appears inside profile context when account state requires it

What is already strong:

  • web auth already handles sign-up, sign-in, recovery, redirect, reauth, and post-sign-out states
  • auth and recovery are integrated with onboarding resume logic
  • mobile auth already understands reauth and session refresh

Redesign liabilities:

  • web auth is more polished and atmospheric than most authenticated shell surfaces, so the product handoff feels uneven
  • mobile auth being embedded in profile is a structural mismatch with both web auth and mobile onboarding needs
  • first impression, account recovery, and signed-in shell posture still read as separate chapters rather than one coherent product

What redesign must preserve:

  • recovery without enumeration
  • redirect-safe re-entry
  • onboarding resume behavior
  • session refresh and sign-out controls

7. Cross-Surface Findings#

The inventory points to seven redesign truths:

  1. The repo already has real product breadth. The redesign problem is not missing functionality on web customer shell routes; it is composition, hierarchy, and continuity.
  2. Home, explore, library, and profile all still default to stacked modules and border-first containers. V1-CSH-063 is justified by current code and visual evidence.
  3. Notifications, activity, and profile each own part of the same “what needs my attention” job. That attention model needs to be redistributed more clearly across surfaces.
  4. Onboarding and auth currently provide a cleaner first impression than most authenticated shell routes. The redesign should raise the shell, not flatten onboarding down.
  5. Mobile parity is incomplete in exactly the places where continuity matters most: auth and onboarding are embedded, while notifications are split between support components and the activity tab.
  6. Shell boundary logic exists in the codebase, but the visual system still lets operator access, domain switching, assistant entry, and notification triage compete too often in the same plane.
  7. The next redesign tasks should preserve existing cross-domain logic and remove visual and interaction noise around it, not restart the information architecture from zero.

8. How This Inventory Drives the Next Tasks#

  • V1-CSH-062: use the cross-surface findings here to define one coherent shell frame across web and mobile
  • V1-CSH-063: remove the card-mosaic default identified across home, explore, library, and profile
  • V1-CSH-064: define one dominant focal plane per route, especially for home, explore, onboarding, and auth
  • V1-CSH-065: redesign navigation, assistant entry, and quick actions so they stop leaking into route content
  • V1-CSH-066: consolidate recommendation posture now that home and explore inventory is explicit
  • V1-CSH-067: redesign route states once each surface has a calmer base composition
  • V1-CSH-068: reduce setup and launch friction using the auth/onboarding/home findings here
  • V1-CSH-069: align assistant, notifications, activity, and library into one persistent-context model
  • V1-CSH-070: validate the resulting redesigns against the explicit web/mobile parity gaps documented above