Description:
I am currently reviewing the A2A Project documentation and have a question regarding the relationship between Agent Servers and Agent instances. Specifically, I would like to know if it is stated or implied that one Agent Server can manage multiple Agent instances.

In the core concepts section, while it is clear that an Agent Server acts as a centralized unit that handles tasks and provides responses, the documentation does not explicitly mention whether multiple Agent instances can be associated with a single Agent Server.

I have looked at several implementations within the A2A ecosystem, and it appears that none of them support this feature of having multiple Agent instances managed by a single Agent Server. It seems that the term "Agent" in this context might actually refer to an "Agent Group," which can contain multiple Agent instances, allowing different tasks to be executed by different Agent instances. Additionally, the Agent Card seems to describe the capabilities of an Agent Group.

Understanding this relationship is crucial for my implementation and design choices. Could you please provide clarification on whether one Agent Server can correspond to multiple Agent instances? Any examples or further details on how this is typically managed in practice would be greatly appreciated.

Thank you for your assistance!

Link: [Documentation Link](https://a2aproject.github.io/A2A/latest/specification/#2-core-concepts-summary)