# 🧠 MCR4 - Neurosymbolic Reasoning Platform
MCR4 combines Large Language Models (LLMs) with symbolic reasoning (Prolog) to create verifiable, explainable AI
systems.
## ✨ Features
- Hybrid reasoning combining neural networks with symbolic logic
- Stateful sessions with persistent knowledge bases
- Real-time WebSocket API (MCP protocol)
- Interactive CLI interface
- Docker container support
- Ontology-aware knowledge validation
- Translation strategies with langchain.js
## 🚀 Quick Start
### Run with Node.js
```bash
# Clone repository
git clone https://github.com/your-repo/mcr4.git
cd mcr4
# Install dependencies
npm install
# Start MCP server
node src/mcp/server.js
# Use CLI interface in another terminal
node src/cli/index.js
```
### Run with Docker
```bash
docker build -t mcr4 .
docker run -p 8080:8080 -e OPENAI_API_KEY=your_api_key mcr4
```
## 📖 CLI Usage
```bash
mcr> new
Created session: session_12345
mcr> assert "Socrates is a man"
Asserted: man(socrates)
mcr> assert "All men are mortal"
Asserted: mortal(X) :- man(X)
mcr> query "Is Socrates mortal?"
Answer: Yes, Socrates is mortal.
```
## 📡 API Reference
Connect via WebSocket: `ws://localhost:8080`
### Message Format
```json
{
"type": "session.create",
"payload": {}
}
```
### Operations:
- `session.create` - Create new session
- `session.assert` - Assert natural language to KB
- `session.query` - Query session with natural language
## 🐳 Docker Deployment
```bash
docker-compose up
```
## 📂 Project Structure
```
mcr4/
├── src/
│ ├── mcr/ # Core reasoning services
│ ├── mcp/ # Model Context Protocol
│ └── cli/ # Command-line interface
├── strategies/ # Translation strategies
├── tests/ # Comprehensive tests
├── Dockerfile # Containerization
└── docker-compose.yml # Development environment
```
## 🧪 Testing
Run tests with:
```bash
npm test
```
## 📜 License
MIT
