root-cause-tracing

Use when errors occur deep in execution and you need to trace back to find the original trigger — systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior