implementing-network-deception-with-honeypots

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The agent parses and analyzes local log files generated by honeypot interactions, which can contain attacker-submitted inputs such as usernames and commands.