Binder integration?

[petebachant]

Binder is a cool way to spin up Jupyter Notebooks from a project that actually have a remote server behind them. Could we possibly inject a Calkit token into the env vars and pull data with DVC to enable interacting with larger datasets?

What other kinds of use cases could that support? One obvious one is for any user to run the DVC pipeline, but sometimes notebooks are treated as pipelines in and of themselves. Is that a supportable practice?