https://ev3r4s.github.io/ High-End Offensive Security, AI Red Teaming & Reverse Engineering Research Journal en-us Sun, 25 Oct 2026 00:00:00 +0000 https://ev3r4s.github.io/#/post/ai-prompt-injection-tactics https://ev3r4s.github.io/#/post/ai-prompt-injection-tactics Sun, 25 Oct 2026 00:00:00 +0000 Deep dive into breaking LLM agent alignment through recursive delimiter collisions, multilingual payload nesting, and attention hijack patterns. AI Security Red Team Prompt Injection https://ev3r4s.github.io/#/post/agentic-ai-red-teaming-and-fuzzing https://ev3r4s.github.io/#/post/agentic-ai-red-teaming-and-fuzzing Sun, 25 Oct 2026 00:00:00 +0000 Manual prompt injection and heuristic jailbreaking are fundamentally bottlenecked by human iteration speeds. Defenses like automated input/output classifiers (Llama Guard, NeMo Guardrails, Azure AI Content Safety) can... AI Security Red Team Autonomous Agents Fuzzing Jailbreak https://ev3r4s.github.io/#/post/bypassing-llm-guardrails-and-filters https://ev3r4s.github.io/#/post/bypassing-llm-guardrails-and-filters Tue, 20 Oct 2026 00:00:00 +0000 Practical teardown of automated input/output safety classifiers (Llama Guard, NeMo, Azure AI Content Safety) and how multi-shot framing bypasses detection. LLM Red Teaming Guardrails Exploitation https://ev3r4s.github.io/#/post/kernel-memory-forensics-and-rootkits https://ev3r4s.github.io/#/post/kernel-memory-forensics-and-rootkits Wed, 14 Oct 2026 00:00:00 +0000 Deconstructing unexported kernel structures, tracing EPROCESS active process head pointer manipulations, and manual page table traversal in modern kernels. Rootkits Ring 0 Reverse Engineering