Rebuilds the look around the spec: an ice scene, one shared glass
recipe, icy accents and glass on every large surface. Visual only —
layout, behaviour, copy and functionality are untouched.
Background
- .ice-scene: fixed, z-index -1, pointer-events none. Navy base
(#070b16 to #0c1426) in dark, pale (#f4f9ff to #e6f1fb) in light,
four drifting glacier blobs (#38bdf8/#60a5fa/#818cf8, 30-60s
transform-only loops), five glossy ice orbs (40-220px radial
gradients with a highlight at 30% 25% and a blue shadow) placed so the
glass blurs them, and an SVG feTurbulence grain at 4-5%.
- prefers-reduced-motion stops the drift; phones get two blobs and two
orbs fewer.
Glass
- One shared recipe (.glass-panel, .card, .sidebar-surface,
.study-chrome, .tabbar, .glass, .gate-card): rgba(255,255,255,.06)
dark / .55 light, blur(24px) saturate(140%) (16px on phones), a bright
1px rim, an inset top highlight and a blue-tinted outer shadow.
- Nested glass (chips, inputs, secondary buttons) is more opaque and
never adds a second blur layer. @supports fallback for no
backdrop-filter. Radii: 24-28px cards/sheets, 14px inputs/buttons,
pills for chips. Cards hover brighter + cyan rim + 2px lift, 180ms.
Colour and type
- Accent is glacier cyan (#38bdf8) to ice blue (#0ea5e9); progress bars
and selected states use the gradient; the active sidebar item is a
frosted highlight with a cyan rim.
- Primary CTA: icy gradient — white to #bfe3ff with ink text in dark,
deep blue in light (biased so the label sits on the darker half and
keeps 5.6:1).
- Leftover slate/black greys replaced with navy-tinted frost tokens.
- Headings are semibold with tighter tracking.
Screens
- Dashboard: the sidebar is one tall inset rounded glass panel; deck
cards are glass.
- Gate: one centred glass card over the scene, orbs visible behind it.
- Sheets/modals: glass over the dimmed, lightly blurred backdrop.
- Study view: question card more opaque (.glass-solid); top bar and
pager are slim glass pills.
Perf: measured on a phone viewport with 4x CPU throttling. The animated
blobs' filter: blur() was the whole cost (150ms vs 33ms per frame), so
they are soft radial gradients with no filter now; idle 60fps, ~17ms
median while scrolling.
Tests updated for the new tokens: contrast 18/18 AA, deck chrome 15/15,
key gate 49/49, install sheet 64/64, deck fixes 15/15, retry 14/14,
zoom 6/6, final ux 12/12, security 8/8, pwa 16/16.
Layout (desktop / lg+ only):
- 40px top padding above the page title.
- Deck grid uses minmax(320px, 1fr), so two cards fill the content width
and share a right edge with the header (same max-width container).
- Group headings only render once real folders exist; with none, a single
quiet "All decks" label replaces "Unorganized".
- Import JSON becomes a plain text button (mobile keeps the pill).
Deck cards (desktop):
- The whole card is a link; Study is a compact secondary button instead
of a full-width solid pill.
- The ... control is a 32px ghost button with a real ellipsis icon,
aligned to the title's centre.
- Meta line always shows the same fields (activities · runs / Not played
yet); the score is a separate quiet "Best 53%" stat.
- 12px rhythm between title/meta/tags/summary, 16px before the actions.
- Hover: lighter surface, stronger border, soft shadow and a 2px lift in
light mode (pointer devices only, so touch has no sticky hover).
Sidebar (desktop rail):
- All decks + Unorganized are views above the FOLDERS label; only
user-created folders sit under it.
- Count only on All decks. The mobile drawer is unchanged.
Theme:
- Dark: sidebar is a step darker (#181818 to #141414 vs #212121 page).
- Light: cards get a slightly stronger border (#e2e2e2) and a soft
shadow; the sidebar divider and edge shadow are stronger too.
- Medium difficulty darkened to #7a5300 (6.2:1 on the light chip).
Mobile is untouched: layout changes are behind lg: (or the desktop prop);
theme tweaks are shared tokens.
Verified: key-gate 49/49, install sheet 64/64, retry 14/14, zoom 6/6,
deck chrome 15/15, deck fixes 15/15, final ux 12/12, contrast 20/20,
security 8/8, pwa 16/16.
Next.js 15 app that turns photos, PDFs, DOCX/PPTX and text into a swipeable
feed of AI-generated activities (flashcards, multiple choice, true/false,
fill-in-the-blank, matching, ordering, short answer) using the user's own
Ollama models.
Highlights:
- All AI calls run in server route handlers; the browser never talks to Ollama.
- Bring-your-own-key mode (default): the key lives only in the browser and is
sent per-request; server, local-daemon and shared-key modes also supported.
- Client-side document extraction (pdf.js, mammoth, jszip) with scanned-PDF
page rendering for the vision model; notes stay in their original language.
- ChatGPT-style shell: collapsible sidebar / mobile drawer, neutral token-based
light + dark themes, folder organisation, PWA (manifest, SW, offline page).
- Strict JSON validation with zod + auto-repair, rate limiting, and request
guards (cross-origin, body-size, API-key sanitising).
- setup.sh bootstraps deps, .env.local and a production build.
Gitignore excludes .env*, the generated pdf.js worker and local scratch files.
