🛡️ BlockPot AI

A smart honeypot system that traps attackers, reads their behavior using AI + NLP, and securely logs everything on the blockchain.

---

❓ What Problem Does It Solve?

• 🔍 Real-time threat detection is challenging for security teams.
• 🧾 Traditional logs can be tampered with or deleted, risking loss of forensic evidence.
• 🧠 Analyzing hacker commands manually is slow and often cryptic.

---

⚙️ What Does BlockPot AI Do?

🪤 1. Deploys a Fake System (Honeypot)

• Simulates a vulnerable server (e.g., SSH or Web login).
• Captures all attacker actions: commands, access attempts, tool downloads.

🗣️ 2. Uses NLP to Understand Behavior

• Extracts meaning from attacker commands using spaCy / Transformers.
• Detects keywords like wget, chmod, root, malware.
• Identifies intentions such as:
  • Installing malware
  • Deleting system logs
  • Stealing credentials

🧠 3. Uses AI for Threat Classification

• A TensorFlow model classifies sessions as:
  • ✅ Safe
  • ⚠️ Suspicious
  • ❌ Malicious

🔐 4. Logs Everything on the Blockchain

• Records:
  • Attacker IP
  • Command
  • Threat level
  • Timestamp
• Stores logs immutably using Ethereum/Hyperledger.
• Ideal for digital forensics, audits, and compliance.

---

🖥️ What Does the User See?

• 🌐 Live feed of incoming attack attempts
• 🌍 Attacker IP & geo-location
• 🧾 NLP-based command summaries (e.g., "attempting to download malware")
• ⚠️ Threat level (Low / Medium / High)
• 🔗 Blockchain verification hash per log

---

💡 In One Line:

A smart trap system that catches hackers, understands what they type using NLP, predicts their intent using ML, and stores it forever using Blockchain.

---

🚀 Technologies Used

Tech	Purpose
🐍 Python	Backend + data processing
🧠 TensorFlow	Threat classification model
📜 NLP (spaCy / Transformers)	Command intent extraction
⛓️ Ethereum / Hyperledger	Immutable log storage
🐳 Docker	Honeypot environment
🌐 React.js + Flask	Dashboard + API

---

📂 Future Enhancements

• 🛠️ Custom rule engine for real-time flagging
• 📊 Threat heatmap visualization
• 🧪 Integrate with VirusTotal API for file analysis
• 📍 GeoIP tracking + dark mode dashboard

---

👨‍💻 Built For

• Hackathons
• Cybersecurity competitions
• AI + Blockchain showcases
• Research & education in cyber forensics

---

🚀 How to Impress Judges / Reviewers

✅ Deploy a Fake SSH Server
Use Cowrie or a minimal SSH trap. Show how it logs every command and sends it to the ML classifier.

✅ Real-time AI Analysis
Display live feedback on:

• Command typed (e.g., wget virus.sh)
• NLP summary: "Trying to download malware"
• ML output: Malicious
• Blockchain hash for proof

✅ Interactive Live Demo
Show a live hacking session using terminal → watch results pop up on your dashboard instantly.

✅ Bonus: Add Custom Rules & Alerts
Auto-flag dangerous commands like rm, chmod, ssh. Optionally trigger email/SMS alerts (can be mocked for demo).

✅ Blockchain Explorer Integration
Show each log's blockchain hash using an explorer. Mention:

“Even if they wipe the system, the evidence remains forever.”

---

🧠 Judges Want... You Give Them...

They Want...	You Deliver...
Real-world impact	SSH honeypot + NLP + Blockchain
ML/AI use	Real-time threat classification
Good UX	Live dashboard with clear logs
Innovation	Honeypot × AI × Blockchain combo

---

📸 Screenshots (Optional)

Add demo screenshots or terminal logs here to visually explain the system.

---

📜 License

MIT © 2025 Blockpot

---