python-debugpy

You can attach debugpy to a running long-lived Python process, such as a Hermes gateway or daemon, to inspect its state at a crash site. This remote debugging approach allows you to debug without restarting the service or disrupting active users.

Post-mortem debugging allows you to inspect program state immediately after an unhandled exception occurs. You can use pdb to drop into an interactive REPL at the exact point of failure to examine local variables and call stacks to identify the root cause.

Yes, you can debug failing pytest tests by dropping into an interactive pdb REPL at the point of failure. This local interactive debugging provides full access to local variables and call stacks to step through code and identify the root cause.

Debugging async and multithreaded Python code requires tailored workarounds for common pitfalls. You can use pdb and debugpy to set breakpoints, step through execution, and inspect call stacks in asyncio handlers and multiprocessing contexts.

Yes, debugpy supports the DAP protocol for remote headless debugging of Python code. You can attach it to running processes or launch processes under a debugger from the start to inspect call stacks and local variables.

Use remote-pdb when print statements and basic logging fail to provide enough context to identify root causes across diverse scenarios. It is ideal for hard-to-reproduce bugs in subprocesses, async handlers, and long-running services.