Darkelf Browser

Post-Quantum | Tor-Enabled | Anti-Forensics | ML Threat Detection

🧬 Quantum-Ready Editions

• ML-KEM-768: Balanced performance & quantum safety
• ML-KEM-1024: Maximum-strength PQ security
• Powered by liboqs with AES-GCM + HKDF session key derivation
• All keys and secrets stored in volatile memory only
• JavaScript API via darkelfCrypto for PQ-safe crypto inside browser context

🧠 JavaScript Threat Detection via ML

• Real-time detection using a trained RandomForestClassifier
• Entropy analysis + eval/WebSocket/fingerprinting API usage detection
• Blocks malicious scripts inline at runtime
• Model integrity verified via SHA256 hash

🔒 Privacy & Anti-Fingerprinting

• Randomized or blocked JS APIs: Canvas, WebGL, Audio, Fonts
• WebRTC, geolocation, timezone spoofing all disabled
• Anti-ETag, spoofed headers, and content security policy (CSP) injection
• No cookies, storage, or disk I/O — ephemeral memory-only architecture

🕵️‍♂️ Anti-Forensics & Stealth

• Disables swap and scrubs memory buffers
• Auto-self-deletion if forensic tools or VMs are detected
• RAM-only logs encrypted with Fernet

🌐 Hardened Networking

• Native Tor integration using stem + SOCKS5 proxy
• Fallback to DoH/DoT if Tor is unavailable
• All traffic padded, delayed, and jittered to evade traffic analysis

📦 Developer Components

• darkelf.py: Main application bootstrapper
• kyber_crypto.py: Kyber-based key encapsulation logic
• PQCryptoAPI: QWebChannel JS-to-C++ bridge for in-browser crypto
• ml_script_classifier.pkl: ML model for JS detection
• scaler.pkl: Normalization scaler for ML inference
• StealthCovertOps: RAM-only stealth logging system
• ObfuscatedEncryptedCookieStore: Ephemeral, encrypted cookie store

📚 Learn More

• GitHub Repository
• Documentation Wiki
• Post-Quantum Editions
• FAQ

📬 Contact

We're open to collaboration, feedback, and research partnerships.