Inspiration
A learner wants to practise reading, but an unreliable internet connection, limited mobile data, or a shared older phone interrupts the lesson. Most learning tools presume that a learner can reconnect, sign in, stream a video, or download a large file. OpenLeaf starts with one narrow problem: can a learner reopen a short story and complete a comprehension activity after losing the signal?
What it does
OpenLeaf — Reading Without a Signal is a responsive, no-login reading classroom. It includes:
- Three original, fictional short reading lessons and comprehension checks with explanations and retries.
- Adjustable text size, a higher-contrast theme, keyboard-operable form controls, and optional Arabic vocabulary hints.
- Optional browser/device text-to-speech where a compatible local voice is available.
- Local completion progress; no learner profiles, analytics, or private student records.
- Teacher Studio: write a short passage and question, then export/import a compact JSON lesson pack for direct sharing.
- A service worker that caches the app after an initial successful HTTPS visit, enabling subsequent use on the same device without a network.
Intended learner and task
A beginning English reader using a basic, older, or shared smartphone needs to read a short story and answer comprehension questions, even when home connectivity is unreliable. A teacher needs to distribute a simple activity without requiring every learner to maintain an online account.
Concrete before and after
Before: A learner opens an online reading exercise at school but loses connectivity on the way home. They cannot reopen the lesson or submit an answer when the network disappears.
After: The learner visits OpenLeaf once while connected and receives the app-shell cache. Later, they reopen a story and complete its comprehension questions offline on the same device. The teacher can share a small, reusable JSON lesson file directly without a login or server.
Try it — live demo
Open: https://openleaf-accessed-hacks.vercel.app/
- Select Start a lesson to open After the Rain.
- Turn on Word hints, increase reading-text size, or switch contrast.
- Answer a question incorrectly to see explanatory feedback, select Need a hint?, and retry.
- Complete all three questions and return to the library to see locally stored progress.
- Enter Teacher studio, create a short fictional reading exercise, download its JSON, and import the same file.
- For the offline test, open OpenLeaf over HTTPS, wait for a ready-offline indication, then turn off connectivity and refresh in the same browser.
No credentials or private information are needed for judging.
How we built it
Semantic HTML, responsive CSS, dependency-free vanilla JavaScript, browser localStorage and Cache API/service worker, optional Web Speech API, and Node.js built-in tests. The lessons, UX, visual system, interactions, and source were produced for this event on October 28, 2026. The source of the standalone application is available through View Page Source at https://openleaf-accessed-hacks.vercel.app/index.html ; offline-cache source is at https://openleaf-accessed-hacks.vercel.app/sw.js . The application is MIT licensed.
Challenges, testing, and limitations
Seven unit tests passed, and browser tests of the standalone build exercised reading, Arabic hints, text size, answering/retrying, completion, teacher creation, JSON import/export, and mobile-width layout. The testing environment restricted browser network navigation, so a full remote-browser E2E test and offline reload were not independently verified there. The hosting platform reports the production deployment as READY.
Offline use requires a first successful visit over HTTPS/localhost, and browsers may evict cache or local data. Text-to-speech quality and offline availability depend on installed device voices. Arabic hints are limited vocabulary assistance, not full translation. Teacher Studio currently authors one-question packs. There has been no formal WCAG audit or learner usability study.
What we learned and what is next
Meaningful accessibility comes from completing a specific task under real constraints—not simply adding an accessibility setting. Next we would conduct keyboard and screen-reader usability tests, test older Android devices and storage eviction, add full Arabic interface localization, and interview learners/teachers on offline workflows.
Credits and pre-existing work
Event contribution: app, code, original demo passages, visual assets, interface and test suite built October 28, 2026, within the Oct 17–Nov 20 development window. No pre-existing foundation reused. No proprietary school data, personal records, licensed photos, or third-party datasets. AI assistance was provided by OpenAI ChatGPT (GPT-6) for planning, implementation, copy, and test preparation; author reviewed and integrated the output. One contributing author: Samer Tallauze.
Built With
- css3
- html5
- javascript
- localstorage
- node.js
- progressive-web-app
- service-worker
- web-speech-api

Log in or sign up for Devpost to join the conversation.