JacketScore

Scope: This README replaces prior selected overview docs

Overview

Web application that helps users decide whether they need a jacket based on weather conditions, location, venue type, arrival time, duration, and personal preferences. Combines real-time weather data with AI-generated recommendations to provide a personalized jacket score (0-100%) and tailored advice.

Live and Admin

• 🌐 App URL: https://jacketscore.com
• 🚀 Vercel Dashboard: Connected to GitHub for continuous deployment
• 🌍 Domain: jacketscore.com (purchased via GoDaddy)
• 📧 Comments: Email notifications via Gmail OAuth2 to YOUR_EMAIL
• 📊 Analytics: Usage tracking and performance monitoring

Tech Stack

• ⚛️ Frontend: Next.js 14 + React 18 + Tailwind CSS
• 🎨 UI: Responsive design with dark mode support
• 🚀 Hosting: Vercel with serverless functions
• 🗺️ Location: Google Places API with autocomplete
• 🌤️ Weather: OpenWeather API (5-day/3-hour forecasts)
• 🤖 AI: OpenAI ChatGPT API for personalized recommendations
• 📧 Email: Nodemailer with Gmail OAuth2 for comment system

Quick Start

git clone https://github.com/bradyespey/jacket-score
cd JacketScore
npm install
npm run dev

Environment

Required environment variables:

# Google Places API
NEXT_PUBLIC_GOOGLE_API_KEY=YOUR_GOOGLE_API_KEY

# Weather Data
OPENWEATHER_API_KEY=YOUR_OPENWEATHER_API_KEY

# AI Recommendations
OPENAI_API_KEY=YOUR_OPENAI_API_KEY

# Email System (Gmail OAuth2)
GMAIL_USER=YOUR_GMAIL_USER
GMAIL_CLIENT_ID=YOUR_GMAIL_CLIENT_ID
GMAIL_CLIENT_SECRET=YOUR_GMAIL_CLIENT_SECRET
GMAIL_REFRESH_TOKEN=YOUR_GMAIL_REFRESH_TOKEN

Run Modes (Debug, Headless, Profiles)

• 🐛 Debug Mode: npm run dev with console logs for API calls and data handling
• 🌐 Production Mode: Deployed on Vercel with environment variables
• 📱 Mobile Mode: Responsive design optimized for mobile devices

Scripts and Ops

• 🔧 Development: npm run dev — Start local development server at localhost:3000
• 🏗️ Build: npm run build — Next.js production build
• 🚀 Start: npm start — Start production server
• 🔍 Lint: npm run lint — ESLint code checking
• 📦 Deploy: Automatic via GitHub integration to Vercel

Deploy

• 🚀 Platform: Vercel connected to GitHub repository
• 📦 Build Command: npm run build
• 📁 Output Directory: .next
• 🌐 Domains: jacketscore.com (primary), vercel app domain
• 🔄 CI/CD: Automatic deployment on GitHub push

App Pages / Routes

• 🏠 Main Page (/): Location input, venue type, arrival time, duration, gender selection with Google Places autocomplete
• 📊 Results: Jacket score display with progress bar, weather information, AI recommendation, and edit inputs option
• 📧 Comments: Integrated comment system with email notifications

🤖 How It Works

1. User Inputs:
   • Select location, venue type, arrival time, duration, and optionally, gender
2. Weather Fetching:
   • Retrieves weather data for the selected time and location
3. AI Recommendation:
   • Sends data to OpenAI API for a personalized recommendation
4. Jacket Score Calculation:
   • Calculates a score from 0-100% indicating the need for a jacket
5. Results Display:
   • Shows the jacket score, AI recommendation, and weather information

Directory Map

JacketScore/
├── src/
│   ├── app/
│   │   ├── api/
│   │   │   ├── calculateJacketScore/    # Jacket score calculation logic
│   │   │   ├── getChatGPTRecommendation/ # OpenAI API integration
│   │   │   ├── getWeather/              # OpenWeather API calls
│   │   │   └── submitComment/           # Email notification system
│   │   ├── page.js                      # Main frontend with state handling
│   │   └── layout.js                    # App layout and global styles
│   └── components/
│       ├── PlaceAutocomplete.js         # Google Places autocomplete with city/state
│       └── JacketScore.js               # Score display component
├── public/images/                       # App logos and screenshots
└── tailwind.config.js                  # Tailwind CSS configuration

Troubleshooting

• 🔑 API Keys: Ensure all environment variables are properly set in Vercel
• 🗺️ Location Issues: Google Places API restricted to localhost and jacketscore.com domains
• 🌤️ Weather Data: OpenWeather API rounds temperature/wind to whole numbers, matches 3-hour intervals
• 🤖 AI Responses: ChatGPT prompts include optional gender, city, state based on user input
• 📧 Email System: Gmail OAuth2 requires proper GCP setup with authorized redirect URIs
• 📱 Mobile UI: Button centering and responsive design optimized for mobile devices

AI Handoff

Read this README, scan the repo, prioritize core functions and env-safe areas, keep env and rules aligned with this file. Focus on weather API integration, AI recommendation logic, and jacket score calculation algorithms.