SYSTEM FEATURES

Deep dive into every component of the NeuralCTF Engine

🎯

Adaptive Simulation Engine

Generates randomized virtual environments including Web Servers, Databases, API Gateways, Cloud Nodes, Workstations, and Firewalls. Choose from predefined scenarios like CyberHeist 2077, Neural Hospital, and Shadow Grid, or let the engine build a unique infrastructure every run.

🤖

AI Defender System

Powered by scikit-learn RandomForestClassifier. The defender trains on synthetic feature vectors (security level, vulnerability count, alert history, attack phase) and predicts compromise probability per node in real time. It dynamically raises authentication thresholds, deploys virtual honeypots, and recommends patches.

🕸️

Interactive Network Graphs

Plotly-powered force-directed topology visualization built on NetworkX. Nodes glow green when secure, yellow when alerted, and red when compromised. Hover for detailed node inspection including open ports, services, and vulnerability status.

⚔️

Safe Attack Chain Simulation

Six fully sandboxed educational phases: Reconnaissance, Scanning, Credential Discovery, Privilege Escalation, Lateral Movement, and Data Exfiltration. No real exploitation code is ever executed. Perfect for CTF training and defensive skill development.

📊

Cyberpunk Dashboard

Dark-mode Streamlit interface with neon green, blue, and red accents. Includes animated metric cards, risk heatmaps, a hacker-style terminal panel, and an AI system monitor. Fully responsive and internship-portfolio ready.

📁

Exportable Reports

Export complete session summaries as JSON with timestamps, metrics, and defender actions. Download node inventories as CSV for spreadsheet analysis. Generate matplotlib status distribution charts for presentations.

🧠

Explainable AI

Understand why the AI Defender makes decisions. Feature importance visualization shows exactly which factors (security level, vulnerability count, alerts, phase index) drive threat predictions and defensive recommendations.

🎮

Difficulty Scaling

Four difficulty tiers: Easy (5 nodes, low vulnerabilities), Medium (8 nodes, balanced), Hard (12 nodes, aggressive AI), and Insane (16 nodes, relentless defense). Each tier adjusts security levels, vulnerability density, and AI aggression.

🔧

Patch Management

Interactive vulnerability patching directly from the Network Graph. Click any discovered vulnerability on any node to remediate it. Watch security levels increase and threat levels decrease in real time.