Nested

Learn a language by diving through its own words.

Look up a word, read its definition — then tap any word inside that definition to dive into its definition, and keep going. Every word is a door. It's the monolingual-immersion method turned into something you can actually play with: you stay inside the target language the whole way down.

🔗 A single static HTML file. No build, no backend, no tracking. Open index.html and it just works.

Core features

• Infinite dive — every word in a definition, example, or synonym list is clickable; tapping it opens that word's own definition.
• Dive trail — a breadcrumb of the path you've taken; tap any step to jump back.
• Pronunciation — real human audio when available, browser speech synthesis as fallback.
• Save words — star a word to keep it; saved words live in your browser (localStorage, capped at 500 words). Nothing leaves your device.

Language support

The concept is language-agnostic. v1 ships with English, because the underlying Free Dictionary API currently only serves reliable data for English. Other languages arrive once a multilingual data source is wired in (see roadmap).

Roadmap / TODO

• More languages — swap in Wiktionary's REST API to genuinely support any language (needs HTML parsing; bigger lift).
• AI assist — optional per-word "explain simply / examples / etymology" via an LLM, behind a backend proxy (no API keys in the browser).
• Sign in with Google — sync saved words across devices (needs a backend / Firebase / Supabase). Local-only for now.
• Export saved words (CSV / Anki).

Run locally

Just open index.html in a browser. Or serve it:

python -m http.server 8000   # then open http://localhost:8000

Deploy

It's a static site — any static host works.

• Vercel: import this repo at vercel.com/new and deploy. vercel.json is included.
• GitHub Pages: enable Pages on the main branch (root) in repo Settings.

License

MIT