GIAC AI Platform Security

A security architect is designing a multi-agent GenAI application and needs to select a communication protocol that allows agents and tools to exchange context. During the design review, the team debates which protocols and standards are relevant to securing agent-to-agent and agent-to-tool communication. Which set of protocols is specifically identified as relevant to agents communication in the course material?

An engineer securing an MCP-based agentic deployment is asked to enumerate the attack techniques that specifically threaten MCP infrastructure so they can build appropriate detections. Based on the course content covering MCP Attacks and OAuth Security, which attacks are called out as specific to MCP?