English Noting

Never forget a word you've learned — and only review what actually matters.

An AI-powered English vocabulary app. Capture words while reading or watching, get AI-generated memory cards instantly, and review only what your brain actually needs using a smart priority score.

What it does

• Capture words in one tap from any context (reading, video, conversation)
• AI memory cards — definition, example sentences, part of speech, CEFR level, generated asynchronously
• Smart review queue — Memory Priority Score (MPS) ranks words by time since last review, accuracy, confidence, failure patterns, and word frequency
• Adaptive quiz formats — new words get multiple choice; well-known words get fill-in-the-blank
• Calendar view — track daily vocabulary streaks and stats

Project structure

English-noting/
├── engnoting-frontend/     # Single-page app (HTML + React via CDN, no build step)
│   ├── Eng-noting.html     # Entry point
│   ├── app.jsx             # Root component, screen routing
│   ├── screens/            # Dashboard, Library, Review, Word Detail, MPS, Settings, Auth
│   ├── sidebar.jsx         # Navigation sidebar
│   ├── capture.jsx         # Floating capture button
│   └── styles.css
│
└── engnoting-backend/      # Go REST API
    ├── cmd/api/main.go     # Entry point
    ├── internal/
    │   ├── domain/         # Core types and business rules
    │   ├── usecase/        # Application logic
    │   ├── http/           # HTTP handlers and middleware
    │   └── infrastructure/ # Database repos, AI clients, JWT
    └── migrations/         # PostgreSQL migrations

Tech stack

Layer	Technology
Frontend	HTML, CSS, React (via CDN), plain JSX
Backend	Go 1.25, chi router
Database	PostgreSQL
Auth	JWT
AI	OpenAI GPT-4o-mini, Google Gemini, DeepSeek

Quick start

Backend

Prerequisites: Go 1.25+, PostgreSQL 12+, migrate CLI

cd engnoting-backend

# Copy and fill in env vars
cp .env.example .env

# Run database migrations
make migrate-up  # requires DATABASE_URL set in env

# Start the server (with live reload via air)
air
# or without live reload:
go run cmd/api/main.go

Environment variables (.env):

DATABASE_URL=postgres://user:password@localhost:5432/engnoting?sslmode=disable
AI_API_KEY=sk-your-openai-api-key
AI_PROVIDER=openai          # openai | gemini
PORT=8080
JWT_SECRET=your-secret
JWT_EXPIRE_MINUTES=60

Start PostgreSQL with Docker:

docker-compose up -d

Frontend

No build step required. Open engnoting-frontend/Eng-noting.html directly in a browser, or serve it with any static file server:

cd engnoting-frontend
npx serve .
# or
python3 -m http.server 3000

Point the frontend API base URL to your running backend (http://localhost:8080).

API overview

All endpoints require Authorization: Bearer <token>.

Method	Path	Description
POST	/api/auth/register	Register a new user
POST	/api/auth/login	Log in, receive JWT
POST	/api/words	Add a new word (AI explanation is async)
GET	/api/words	List all words
GET	/api/words/{id}	Get word detail
POST	/api/v1/reviews/session	Start a review session
GET	/api/v1/reviews/session/current	Get current session item
POST	/api/v1/reviews/submit	Submit a review result
POST	/api/v1/reviews/session/advance	Skip current item
GET	/health	Health check

Full API docs with request/response examples: engnoting-backend/README.md

How MPS works

The Memory Priority Score (0–100) is calculated deterministically:

MPS = (time_factor × 30) + (accuracy_factor × 30) +
      (confidence_factor × 15) + (failure_factor × 15) +
      (frequency_factor × 10)

Same inputs always produce the same score — no black-box ML. Users can see exactly why each word is queued for review.

Running tests

cd engnoting-backend
go test ./...

License

MIT