Learn how Datadog AI Guard can flag self-replicating prompt injections when an AI agent copies hidden instructions into email replies or skill files.