What problem does it solve?
It helps you diagnose confusing Python failures by letting you inspect variables, step through code, and debug long-running processes that you cannot easily restart.
Core Features & Use Cases
- Break at the exact line with
breakpoint() or python -m pdb for interactive local inspection when a traceback alone is not enough.
- Remote/Headless debugging with debugpy (DAP) to attach to already-running processes and debug Hermes-style gateway, daemon, or child workers.
- Post-mortem inspection to open a debugger session at the crash site and examine locals in the failing frame.
Use case: Your Hermes gateway misbehaves and you only have symptoms; attach a DAP client to the live process, set a breakpoint in the handler code, and step through the failing logic.
Quick Start
Run your failing service with debugpy listening and wait for a client attachment so you can set breakpoints and inspect locals in the moment.