Same channel cannot be reused for more than one execute command per SSH protocol. There are no examples because it is not possible.

The channel object is available on all clients as well as output objects, so clients can call channel.shell() and write commands manually to an interactive shell. This will not give separate output streams, that also needs to be handled in code. Providing an example of this is literally implementing this feature. PR welcome.

Regarding execute test, it is calling run_command which calls execute and that is what is being tested. What do you suggest the test should be doing?

The public API is fully documented and most developers are able to extend functionality when needed based on that information. Suggestions and PRs are always welcome.

The tests are for testing implementation, not as user code examples.

There is function to open a channel, which execute takes as param. I'd expect most developers to be able to work with the documented API as-is, without specific examples for every function. PR welcome if you feel a specific example would be useful. Any further on this should be on the mail group, this is an issue for tracking feature implementation.

See Code of Conduct. Abuse will not be tolerated.

As with all open source projects, work is done on a volunteer basis by people offering their free time. If there is something missing that someone wants and is sufficiently motivated to do they are more than welcome to.

All open source projects will say the same thing. Want something done quickly - roll up your sleeves and get going. Expect work for free and tell people with abusive language what to do.. good luck with that.