Methodologies & posture

All offensive work is contracted and authorised. Operators use callsigns only (Ivxx–Vxx). Zero-knowledge report storage and crypto/XMR billing available for sensitive private engagements.

AI & LLM Security

Authorised assessments only. Target persona: AI Startups, Enterprise AI/LLM Security, FinTech.

• Prompt injection and jailbreak assessments
• Model poisoning and training-data extraction risks
• Agentic workflow exploitation
• RAG-pipeline poisoning audits