What happens if a hypothesis is proven wrong?

The skill marks the hypothesis as disproven, adds an 'Iterate' task to the workflow, and requires the formation of a new, evidence-based hypothesis before proceeding.

When should I use the 'find-root-causes' skill instead?

Use 'find-root-causes' for formal incidents that require full Root Cause Analysis (RCA) documentation; it utilizes this debugging skill as its underlying investigative engine.

Is a failing test always required?

Yes. The skill's 'Iron Law' requires creating a test that reproduces the bug before any fix is implemented to ensure the solution is verified and prevents future regressions.

Does this skill work for performance issues?

Yes, it is specifically designed to handle performance bottlenecks, memory leaks, and high CPU usage by treating them as systematic defects requiring evidence collection.

How does the Systematic Debugging skill differ from standard bug fixing?

It enforces a mandatory four-stage investigation framework that prohibits proposing solutions until a root cause is proven through evidence, preventing 'guess-and-check' coding.

Systematic Debugging

Name: Systematic Debugging
Author: outfitter-dev

byoutfitter-dev

•

보안 및 테스팅

Resolves complex software defects using an evidence-based, four-stage framework that prioritizes root cause investigation over trial-and-error fixing.

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.

주요 기능

01Automated stage tracking and progress management

02Evidence-based four-stage debugging framework (Investigate, Isolate, Hypothesize, Verify)

03Strict 'No Fix Without Root Cause' enforcement policy

04Mandatory failing test creation prior to implementation

0518 GitHub stars

06Circuit breaker logic to escalate after three failed hypotheses

사용 사례

01Diagnosing performance regressions such as memory leaks or high CPU usage

02Troubleshooting intermittent race conditions or timing-dependent failures

03Resolving complex integration failures between APIs and database services

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add outfitter-dev/agents debugging

For use in Claude.ai and ChatGPT

Download Skill