SignalMine – Self-Healing LP Chat

A multi-agent system that converts natural language optimization problems into validated Linear Programs using Pydantic AI agents with self-healing capabilities.

Features

• 🤖 Pydantic AI Agents - Generate Linear Programs from natural language
• 🔧 Self-Healing - If LP validation fails, a fixer agent automatically corrects it
• 🎨 ChatGPT-style UI - Clean HTML/CSS/JS frontend with Bootstrap modals
• 🔄 Model Selection - Switch between GPT-4o, GPT-4o-mini, GPT-4-turbo, GPT-3.5-turbo
• 📝 LaTeX & Python Output - Get mathematical formulations and working code
• 🔐 User Authentication - Register/login with persistent sessions
• 💬 Chat History - Save and load previous conversations
• 🔑 Custom API Keys - Add your own API keys for OpenAI, Anthropic, Google, Groq, and more
• 📊 Usage Logging - Track model communication, tokens used, and response times
• 🗄️ PostgreSQL Database - Neon PostgreSQL for data persistence
• ☁️ Vercel Ready - Deploy to Vercel with serverless functions

---

Database Schema (ERD)

The application uses PostgreSQL with the following tables:

• Users - User accounts (ID, username, password)
• Chat - Chat conversations (ID, userId, Name, originalPrompt, lastMessageId)
• Messages - Chat messages (ID, chatID, message, order, origin)
• Models - User's custom API keys (ID, userId, Name, API-key, provider, base_url)
• Logs - Model communication logs (ID, messageId, log, model_used, tokens_used, response_time_ms, was_healed)

---

Deploy to Vercel

1. Push to GitHub

git add .
git commit -m "Add database and auth support"
git push

2. Set up Vercel Environment Variables

Add these environment variables in Vercel dashboard:

OPENAI_API_KEY=sk-your-key-here
PGHOST=your-neon-host.neon.tech
PGDATABASE=neondb
PGUSER=neondb_owner
PGPASSWORD=your-password
PGSSLMODE=require
SECRET_KEY=your-secret-key

3. Deploy on Vercel

1. Go to vercel.com and import your GitHub repo
2. Add the environment variables above
3. Click Deploy

---

Repository Scope

This repository is organized around the Vercel deployment path only:

• api/ contains the serverless Python handlers
• public/ contains the static frontend assets served by Vercel
• vercel.json defines routing between the frontend and API handlers
• requirements.txt and api/requirements.txt define Python dependencies for deployment

The older local Flask server path has been removed to keep the workspace aligned with production.

---

API Endpoints

Authentication

• POST /api/auth - Login or register (action: 'login' or 'register')
• GET /api/auth - Validate token and get user info
• DELETE /api/auth - Logout

Chat

• POST /api/chat - Generate LP from prompt
• GET /api/chats - Get user's chat history
• GET /api/chats/:id - Get specific chat with messages
• POST /api/chats - Create new chat
• DELETE /api/chats/:id - Delete chat

Models

• GET /api/models - Get available default models
• GET /api/user-models - Get user's custom models
• POST /api/user-models - Add custom model/API key
• PUT /api/user-models/:id - Update custom model
• DELETE /api/user-models/:id - Delete custom model

Logs

• GET /api/logs - Get user's usage logs with summary

---

Custom API Keys

Users can add their own API keys for various providers:

Provider	Base URL
OpenAI	https://api.openai.com/v1
Anthropic	https://api.anthropic.com/v1
Google AI	https://generativelanguage.googleapis.com/v1beta
Groq	https://api.groq.com/openai/v1
Together AI	https://api.together.xyz/v1
Custom	User-provided

---

Project Structure

Signal-Mine-Senior-Project/
├── api/
│   ├── auth.py         # Authentication API
│   ├── chat.py         # LP generation API
│   ├── chats.py        # Chat management API
│   ├── database.py     # Database connection & models
│   ├── health.py       # Health check API
│   ├── logs.py         # Usage logs API
│   ├── models.py       # Available models API
│   ├── user_models.py  # Custom API keys API
│   └── requirements.txt
├── public/
│   ├── app.js          # Frontend JavaScript
│   ├── index.html      # Main HTML page
│   └── styles.css      # Styles
├── .env                # Environment variables
├── requirements.txt    # Python dependencies
└── vercel.json         # Vercel configuration

---

License

MIT