Skip to content

README.md

Latest commit

 

History

History
109 lines (78 loc) · 4.21 KB

README.md

File metadata and controls

109 lines (78 loc) · 4.21 KB
Raw

DECODED — Dating Intelligence OS

Next.js 14 TypeScript Groq Cloud Tailwind CSS Framer Motion License: MIT

An elite behavioral psychology & texting subtext analyzer engineered with radical honesty, non-neediness principles, and ultra-fast inference.

⚡ Core Systems

1. Voice Calibration Engine (Anti-Impostor Voiceprint)

  • Calibrate your natural texting cadence with toggles for All-Lowercase, Dry / Deadpan, Fast & Punchy (< 7 words), and Zero Emoji.
  • Pass custom texting samples to mathematically constrain all suggested replies to sound unmistakably like you.

2. Contact Dossiers & Trajectory Tracking

  • Switch between active contact profiles (Sarah (Hinge), Alex (Bumble), Quick Scan).
  • Record multi-session conversation history and track momentum across scans:
    • 📈 Accelerating Interest (Frame holding / high engagement)
    • ⚖️ Stable / Plateau (Balanced frame)
    • 📉 Decelerating / Frame Loss (Fading / chasing dynamic)

3. The "Walk-Away" Dignity Diagnostic

  • Automatically detects severe effort mismatch, dry responses, or fading dynamics.
  • Delivers a dedicated Walk-Away Play with explicit dignity rules and re-engagement triggers.

4. High-EQ Behavioral Psychology Engine

  • Built on radical honesty, non-neediness, and outcome independence ("Fuck Yes or No").
  • Enforces authentic human syntax: strictly 5–12 words, zero emojis, and zero pickup artist clichés.

🏗️ Architecture & Tech Stack

decoded/
├── app/
│   ├── api/analyze/route.ts   # Edge runtime LLM reasoning engine
│   ├── globals.css            # Dark frosted-glass design tokens & animations
│   ├── layout.tsx             # Standalone PWA shell & typography
│   └── page.tsx               # Reactive dashboard, modals & spring physics
├── lib/
│   ├── types.ts               # Data models & schemas
│   └── utils.ts               # Helper utilities
└── public/
    ├── icon-192.png           # PWA standalone application icon
    ├── icon-512.png           # High-resolution splash icon
    ├── apple-touch-icon.png   # iOS home screen icon
    └── manifest.webmanifest   # PWA manifest

🔐 Environment Variables

Configure the following variables in your .env.local or hosting provider:

Variable Recommended Value Description
LLM_API_KEY gsk_... Groq Cloud API Key
LLM_BASE_URL https://api.groq.com/openai/v1 OpenAI-compatible endpoint
LLM_MODEL llama-3.3-70b-versatile Ultra-fast high-capacity model

Note: If no API key is provided, the application runs seamlessly using its built-in rule heuristics engine.

💻 Local Setup

# Clone the repository
git clone https://github.com/comerade2134/decoded.git
cd decoded

# Install dependencies
npm install

# Run local development server
npm run dev

# Run production build validation
npm run build

🚀 1-Click Deployment (Vercel)

  1. Push this repository to GitHub.
  2. Import into Vercel.
  3. Add your LLM_API_KEY, LLM_BASE_URL, and LLM_MODEL under Environment Variables.
  4. Click Deploy.

🛡️ Privacy & Local Storage Guarantee

  • Zero Database Persistence: All contact dossiers, scan histories, and voice profiles are stored exclusively in your browser's localStorage.
  • Zero Tracking: Your personal conversations are never logged, stored on external servers, or used for model training.

📄 License

Distributed under the MIT License.