🤝 Top Collaborators
🤓 Latest Submissions
PhantomGrid
PhantomGrid is an AI-inspired cyber deception agent built for the AMD Developer Hackathon ACT II. It helps users analyze suspicious cyber inputs such as phishing emails, risky commands, suspicious URLs, and brute-force login attempts. The app takes a suspicious input, classifies the possible threat, assigns a Risk Score, explains why the input is risky, and generates a safe decoy/honeypot-style log. Instead of only detecting or blocking suspicious activity, PhantomGrid demonstrates how deception-based defense can misdirect attackers while keeping real assets protected. The current prototype is built using Python and Streamlit, with rule-based threat detection using curated regex patterns. It includes a live deployed demo, detection history dashboard, decoy log generation, GitHub documentation, and demo screenshots. The project is designed as a beginner-friendly cybersecurity tool, SOC training simulation, and proof-of-concept for deception-based cyber defense. Future improvements include LLM-based explanation mode, AMD GPU/cloud inference, MITRE ATT&CK mapping, file/log upload analysis, and exportable threat reports.
13 Jul 2026
