Prompt Injection covers attacks that manipulate LLMs and AI agents through malicious instructions. Explore direct and indirect prompt injection, jailbreaks, system prompt leakage, poisoned external content, guardrails, detection techniques and practical prevention strategies.