Systematic Debugging

关于

The Systematic Debugging skill transforms Claude into a disciplined engineer by enforcing a rigorous methodology for troubleshooting. It moves through four distinct stages: evidence collection, variable isolation, hypothesis testing, and verified implementation. By requiring a failing test case before any code changes are made and prohibiting 'quick fixes' without a proven root cause, this skill is ideal for tackling intermittent failures, performance bottlenecks, and regressions in complex codebases. It includes built-in task tracking and a circuit breaker to prevent unproductive looping on incorrect assumptions.

主要功能

• Automated stage tracking and progress management
• Evidence-based four-stage debugging framework (Investigate, Isolate, Hypothesize, Verify)
• Strict 'No Fix Without Root Cause' enforcement policy
• Mandatory failing test creation prior to implementation
• 18 GitHub stars
• Circuit breaker logic to escalate after three failed hypotheses

使用场景

• Diagnosing performance regressions such as memory leaks or high CPU usage
• Troubleshooting intermittent race conditions or timing-dependent failures
• Resolving complex integration failures between APIs and database services