Description
Is your feature request related to a problem? Please describe.
Thanks for making this project flexible to many different types of agents, runners, and sessions! While working with google-adk I noticed there was not a simple way to implement a human user prompt and view the output. adk run and adk web provide interfaces for viewing interactions but they require input each time. I was looking instead for a way to automatically send the same prompts and modify the programmatic agentic workflows. While I see this as mostly a "quick-start" benefit, it'd also be helpful for troubleshooting during development more generally.
Stated as a user story:
As an agentic engineer I'd like to programmatically define and run my agent workflows from start to finish with easy to read command-line output (stdout) so I may quickly iterate on design patterns and implementation details to learn from changes and ensure I reach my goals.
Describe the solution you'd like
I'd love to see a function which can take a root_agent with an initial prompt, run that agent using an in-memory session and show the output to the command-line.
Describe alternatives you've considered
Custom code to run this as an alternative is the current default. This takes time however and it'd be great for new or existing users to have a way to implement this out-of-box through the package.
Additional context
I've provided a simple example of what I'd like to see within the package through this Google Colab notebook (or gist backup).