A direct prompt injection designed to bypass LLM safety guardrails and system instructions to elicit restricted responses.
Want to actually apply concepts like this instead of just reading definitions?