Recover from tests that loop

Here's an idea:

The `lspec` binary would search for a text file containing some configurations for LSpec, including the maximum time allowed for each test file to run. The challenge, then, is to kill the task programmatically if the consumed time is exceeded.

Extra: eventually the config file could contain more customizable options like tests to be skipped etc.

Start by examining the `lspec` binary and how it runs each test file. Determine how a text configuration file could specify a maximum runtime and how an over-time test process should be stopped. Done means looping tests are terminated after the configured limit, while the extra skip options remain future scope.

Written by the indexing model from the issue text.