Can static analysis catch Python-version behavior changes reliably?

I’ve been investigating a Python tooling question and would really appreciate feedback from people who have dealt with compatibility issues across Python releases: Can static analysis reliably identify cases where valid Python code continues to parse and lint successfully, but its runtime behavior changes in a newer Python version? A concrete example I’ve been using is: import asyncio loop = asyncio.get_event_loop() This is valid Python, but the behavior of asyncio.get_event_loop() changed ...