{"repo":"Delta-Sec/StegX","free":true,"listed":false,"github":"https://github.com/Delta-Sec/StegX","clone":"git clone https://github.com/Delta-Sec/StegX.git","description":"‎StegX is a modern, open-source steganography toolkit designed for digital forensics experts and cybersecurity professionals. It offers fast, accurate, and stealthy methods by non-linear lsb algorithm to embed and extract hidden data within images outperforming traditional tools, detection evasion, and cross-format compatibility","language":"Python","stars":29,"topics":["aes-256-gcm","anti-forensic","cli","cryptography","cybersecurity","data-hiding","digital-forensics","encryption","image-processing","lsb-steganography"],"license":"MIT","category":"security-tools","readme_excerpt":"StegX 2.0: Authenticated Non-Linear LSB Steganography with Argon2id + AES-GCM StegX hides files inside PNG images using password-shuffled LSB embedding and authenticated encryption. Version 2 is a ground-up security rewrite built around a versioned container format, Argon2id key derivation, domain-separated HKDF sub-keys, LSB-matching (±1) embedding that defeats chi-square steganalysis and — optionally — ChaCha20-Poly1305 dual-cipher, F5-style matrix embedding, adaptive cost-map filtering, keyfile 2FA, plausible-deniability decoy payloads, and k-of-n Shamir secret sharing across multiple cover images. Technical Evaluation Report 📊 For a complete, in-depth evaluation of StegX v2.0's cryptographic strength, statistical invisibility (Chi-Square/Entropy metrics), and a head-to-head performance comparison against legacy tools like Steghide , please read our comprehensive StegX v2.0 Technical Evaluation & Benchmark Report . Steganalysis Resistance 🛡️ StegX v2.0 was tested against 10 detection methods across 4 embedding modes — all returned UNDETECTED . Full methodology and results: Test Cases & Validation Wiki . Category Methods Result ---------- --------- -------- Classical Attacks Chi-Square (p=1.0), RS Analysis, Sample Pair, Entropy ✅ UNDETECTED Image Quality PSNR (72-74 dB), SSIM (0.999998), KL Divergence (10⁻⁶) ✅ IMPERCEPTIBLE Machine Learning Random Forest + Gradient Boosting (SRM features, GroupKFold) ✅ 50% accuracy (random) Deep Learning SRNet CNN — 11.5M params, 60 epoch","default_branch":null,"files":null,"tree":[],"storefront":"/r/Delta-Sec","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/Delta-Sec/StegX/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}