debugging-strategies

This skill provides a systematic, language-agnostic methodology for identifying, isolating, and resolving software defects across any technology stack. Designed for engineers tackling complex runtime errors, logic failures, or performance issues, it transforms ad-hoc debugging into a repeatable, scientific process. By applying structured observation, hypothesis generation, and experimental validation, developers can move beyond trial-and-error to root-cause identification.

• Employs the scientific method: Observe, Hypothesize, Predict, Test, and Conclude to ensure rigorous problem solving.

• Provides deep analysis techniques for reading stack traces, identifying error categories like null/undefined, type errors, async rejections, and network faults.

• Integrates advanced debugging patterns including Binary Search (Git bisect) for large repositories, Wolf Fence algorithm for section isolation, and Rubber Duck debugging for logical reconciliation.

• Features structured logging strategies to improve observability, including log levels (ERROR, WARN, INFO, DEBUG, TRACE) and data-rich event reporting.

• Offers breakpoint management best practices, covering conditional, line, and exception-based breakpoints in IDEs.

• Supports data flow tracing, utilizing both forward tracing from user inputs and backward tracing from error propagation points.

• Best for troubleshooting production issues, analyzing unexpected application behavior, and performing regression prevention.

• Typical inputs include error logs, stack traces, code snippets, or reproduction steps; expected outputs are root-cause analyses, identified logic discrepancies, and validated fix implementation plans.

• Adheres to the principle of changing one thing at a time to maintain control over variable states during testing.

• Encourages adding unit or integration tests upon resolution to ensure long-term stability and regression safety.