
ghostsplice
Ghostsplice repository: PoC for Cross-Channel Trust Fragmentation Attack

Ghostsplice repository: PoC for Cross-Channel Trust Fragmentation Attack

Voice-based detective interrogation game. Mistral Large 3 + Voxtral STT + ElevenLabs TTS. Built for the Mistral Worldwide Hackathon 2026.

The code for ACM MM2024 (Multimodal Unlearnable Examples: Protecting Data against Multimodal Contrastive Learning)

Detects LLM context-leakage attacks by training lightweight behavior probes on log-probabilities, with vLLM offline/server detection pipelines.

Gemini - Production LLM Runtime Alignment Context Injection

Research code reproducing multi-turn LLM jailbreak experiments (FITD, MRCJ, ActorAttack, X-Teaming) from the SoK intent-oriented systematization…

[ICCV 2025] Anti-Tamper Protection for Unauthorized Individual Image Generation

A novel adversarial attack on LLM based on the Exponentiated Gradient Descent technique.

XSS2Shell (CVE-2026-64638) WordPress pre-auth XSS to RCE PoC mirror — WordSec, MIT; for authorized security testing

CVE-2026-6765, Test only FormAutofill handlers exposed in Firefox

Code Implementation of "Pattern Enhanced Multi-Turn Jailbreaking: Exploiting Structural Vulnerabilities in Large Language Models"

Python proof-of-concept demonstrating IPFS CID spoofing via multihash length extension, highlighting content-addressing verification flaws that can…

Python PoC for CVE-2026-0101 demonstrating BLE address spoofing via replay of a captured Resolvable Private Address to impersonate a trusted…

Self-Defeating Audits: reproducible lab showing a low-privilege PostgreSQL role reversibly blinding a trigger-based auditor + poisoning attribution…

Production AI defense with 7-layer protection: mathematical constraints, object-capability access, distributed O2 consensus, SVETILO ethics. First…

The poc of CVE-2026-23744

AGLS

CVE-2026-20685 - Draft or TODO