Test your understanding of insecure inter-agent communication.
1. An attacker who has captured a valid "approve deployment" message from last month replays it today. The deployment agent accepts it. Which property of the communication was missing?
2. A deployment agent receives a message over an authenticated mTLS connection from a verified orchestrator identity. The message content, however, was altered by a compromised logging sidecar that had write access to the Redis queue between the orchestrator and the deployment agent. Which mitigation would have caught this — and which would not?