One article a day, built from the words you are learning.
Save the words you meet. Hear them in a story. Read, then recall.
english.goodone.si · Development · License
Today |
Lesson |
You |
Daily English is a mobile-first PWA for learners who already collect vocabulary and need somewhere to use it. Each lesson is one short article written from the words on your list, read aloud in General American, then checked with a quick recall round. Five minutes, once a day.
Live at english.goodone.si. Sign in with Google.
- Listen — hear the article once, without reading first.
- Read — follow the text, tap a word for a gloss, tap a sentence to hear it again.
- Listen again — the whole article, a little faster, now that you know what it says.
- Recall — a short quiz on the language that actually appeared.
Playback is 0.75× / 1× / 1.25×. Articles stay on your account: an unfinished lesson opens where you left it, and a finished one can be stepped through again.
Between lessons there is Explore, a deck of word cards at your level to flick through — your own saved words, new ones from the pool, or both.
| Piece | Choice |
|---|---|
| App | TanStack Start (React 19) on Cloudflare Workers |
| Data | D1 (SQLite) via Drizzle, R2 for audio |
| Writing | DeepSeek |
| Speech | OpenAI gpt-4o-mini-tts |
| Words | A CEFR-tagged pool in-repo, plus a Chrome extension to save from any page |
The Chrome extension lives in packages/chrome-extension. It adds a selected word to the same list the site uses.
pnpm only. Node 22.
pnpm install
cp .dev.vars.example .dev.vars # then fill in keys you actually need
pnpm mock:ai # http://127.0.0.1:8799 — keep this running
pnpm dev # http://localhost:4000pnpm mock:ai stands in for DeepSeek and OpenAI speech so a local run does not spend API credit. The mock writes placeholder prose on purpose. On the login screen, Sign in as the test learner exists only in development.
pnpm check # typecheck, tests, production build — this is what CI runsMore detail: local development, testing.
src/data/vocabulary.ts is generated from the CEFR-J and Octanove vocabulary profiles (CC BY-SA 4.0, via Open Language Profiles) plus a web-frequency list. Rebuild with pnpm vocab. Do not hand-edit the output.
That file is an adaptation of CC BY-SA 4.0 data, so it stays under CC BY-SA 4.0 whatever the rest of the repository is licensed as. The share-alike condition came with the data and does not go away.
PolyForm Noncommercial 1.0.0. Read it, learn from it, change it, build on it — for any noncommercial purpose. Commercial use is not granted, and that includes running it as a service for other people.
It is not a self-hosting kit either. wrangler.jsonc is wired to the live account, and what actually makes the app work — the domain, the D1 database, the R2 bucket, the model keys — is not in here. There are no instructions for standing up your own copy, and no support for it.


