What problem does it solve?
This Skill helps you diagnose failing Python code by letting you pause execution, inspect state, and step through the exact moment things go wrong.
Core Features & Use Cases
- Interactive breakpoints with pdb: Use breakpoint() for the simplest local REPL at the line where the bug is suspected, including stepping, inspecting locals, and conditional breakpoints.
- Remote/headless debugging with debugpy (DAP): Attach from an IDE or a lightweight client to long-lived processes such as gateways or daemons without restarting them.
- Targeted debugging workflows: Debug pytest failures, run scripts under pdb without source edits, and perform post-mortem inspection of exceptions to see locals at the crash site.
Example use case: your Titan-related test fails but the traceback doesn’t explain why a value is wrong, so you drop into pdb at the failing test frame to inspect locals and control flow, then switch to debugpy when the issue is in a long-running gateway process that can’t be easily restarted.
Quick Start
Use breakpoint() at the suspect line, run the code, and interact at the (Pdb) prompt to inspect locals and step through the failing path.