A team wants a standardized way for an AI agent to discover and use capabilities exposed by external tools and systems. In this context, what is the role of Model Context Protocol (MCP)?

1. It is a training method that permanently writes every tool response into the foundation-model weights.
2. It defines a standardized client-server interaction for exposing capabilities such as tools and contextual resources to AI applications.
3. It guarantees that all tool outputs are factually correct and immune to prompt injection.
4. It is an encryption algorithm that automatically makes every connected tool safe to execute.