feature: CI to make sure that changes don't break A2A behavior

[esnible]

Feature Description

Currently we get a lot of automated dependency update PRs -- updates that have never been fully tested.

It would be good to verify those PRs don't break runtime behavior.

Proposed Solution

Use an a2a CLI with a hard-coded question and perhaps a regex to recognize the answer. Run that automation as part of CI.

Want to contribute?

• I would like to work on this issue.

Additional Context

No response