- API-key gate is now a fixed, out-of-flow overlay (inset-0, overflow
hidden, overscroll-none) with its own internal scroller, so the document
has no scrollable height on phones. It can't be panned or bounced.
- Home/create mains are flex columns with min-h-dvh and a flex-1 content
area, and the empty state's mobile padding was trimmed, so a short/empty
library is exactly viewport height instead of overflowing by ~76px.
Long pages (the create form, decks) still scroll normally.
- html/body get overscroll-behavior: none to stop pull-to-refresh and
rubber-banding on fitting screens.
Verified with a scroll probe (gate and empty library: scrollHeight ==
viewport; create still scrolls) plus zoom 6/6, install sheet 64/64,
pwa 16/16, final-ux 12/12, retry 14/14, security 8/8, contrast 20/20.
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.
