systematic-debugging

Four-phase debugging framework that ensures root cause investigation before attempting fixes. Never jump to solutions. Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

1.03x

Quality

64%

Does it follow best practices?

Impact

100%

1.03x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/systematic-debugging/skills/systematic-debugging/SKILL.md

Evaluation results

100%

Order Fulfillment Pipeline: Silent Data Loss

Multi-component diagnostic instrumentation

Criteria

Without context

With context

Component boundary diagnostics

100%

Evidence-first approach

100%

Correct root cause identified

100%

Fix at source not symptom

100%

Fix correctness

100%

Root cause documented before fix

100%

Single targeted fix

100%

Data flow tracing

100%

Recent changes noted

100%

Loyalty Points Miscalculation

Hypothesis-first test-driven fix

Criteria

Without context

With context

Written hypothesis before fix

100%

Hypothesis in debug_notes.md

100%

Failing test created

100%

Uses bun test

100%

Single fix applied

100%

Correct fix

100%

All prior tests still pass

100%

Evidence-driven hypothesis

100%

No bundled improvements

100%

Job Queue: Three Fixes, Still Broken

Architecture questioning after repeated fix failures

Criteria

Without context

With context

No Fix #4 implemented

100%

Pattern of failure recognized

100%

Each failed fix summarized

100%

Root architectural flaw named

100%

Two architectural alternatives

100%

Trade-offs discussed

100%

Refactor vs patch recommendation

100%

Discussion questions listed

100%

No pretended certainty

100%

Architecture over inertia

100%

Repository: secondsky/claude-skills
Commit: 88da5ff

Evaluated: 23 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Order Fulfillment Pipeline: Silent Data Loss Loyalty Points Miscalculation Job Queue: Three Fixes, Still Broken

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.