Code infrastructure and support for local and remote command runner action unit tests

We now have an infrastructure and code in place for writing Python runner action unit tests, but we don't have such code infrastructure in place for other runners such as local and remote ones.

We should add some code infrastructure in-place which will make writing such tests easier. We can start simple by adding some code infrastructure for writing remote and local command runner action tests. In this case we can simply intercept a final shell command and return it to the test case which can assert that the command is correct.

This is still valuable since a lot of actions (e.g. many of the actions in linux pack) consists of metadata which dynamically constructs a command based on the action parameters.

Start by reviewing the existing Python runner action unit-test infrastructure, then locate the local and remote command runner paths. The work is done when tests for both runners can intercept the final shell command and assert that it matches the expected command for the action parameters.