debugging-wizard

Use when investigating errors, analyzing stack traces, or finding root causes of unexpected behavior. Invoke for error investigation, troubleshooting, log analysis, root cause analysis.

Install with Tessl CLI

npx tessl i github:jeffallan/claude-skills --skill debugging-wizard

What are skills?

Review — 67%

Does it follow best practices?

If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:

npx tessl skill review --optimize ./path/to/skill

Learn more

Evaluation — 98%

Agent success when using this skill

Validation — 11 / 11 Passed

Validation for skill structure

SKILL.md

Review

Evals

Evaluation results

99%

Investigate Intermittent Batch Processing Failure

Structured debug output and root cause discipline

Criteria

Without context

With context

Root Cause section

100%

Evidence section

83%

100%

Fix section

100%

Prevention section

50%

100%

No premature fix

100%

Traces data flow

100%

Regression test written

100%

Test is a failing test first

75%

100%

No debug code in corrected script

100%

Single focused fix

100%

90%

Without context: $0.3795 · 1m 36s · 15 turns · 16 in / 5,630 out tokens

With context: $0.5856 · 2m 40s · 21 turns · 20 in / 7,787 out tokens

96%

12%

Debug a Broken Cart Summary Component

Pattern analysis and hypothesis-driven debugging

Criteria

Without context

With context

Comparison table

66%

83%

Written hypothesis

33%

100%

Root cause identified

90%

100%

Data flow traced

87%

100%

Reproduction pattern noted

100%

Single minimal fix

100%

75%

Fix uses null guard or optional chaining

100%

Test covers undefined cart

100%

Test covers loading state

66%

100%

Root cause before fix

100%

No debug code in fixed file

100%

Without context: $0.3038 · 1m 34s · 11 turns · 12 in / 4,978 out tokens

With context: $0.3578 · 1m 44s · 15 turns · 344 in / 5,501 out tokens

100%

Investigate a Persistent Regression in the Authentication Service

Three-fix threshold and git bisect regression strategy

Criteria

Without context

With context

Bisect plan present

100%

Correct good commit

100%

Automated bisect

25%

100%

Suspicious commits identified

100%

Three failed patches acknowledged

100%

Architectural problem identified

100%

Structural fix proposed

100%

No symptom patch proposed

100%

Pattern of failures documented

100%

Delta debug cited

87%

100%

No debug code in outputs

100%

Without context: $0.2660 · 1m 37s · 8 turns · 9 in / 5,075 out tokens

With context: $0.5052 · 2m 23s · 17 turns · 50 in / 7,588 out tokens

Evaluated: about 6 hours ago
Agent: Claude Code

Table of Contents

Investigate Intermittent Batch Processing Failure Debug a Broken Cart Summary Component Investigate a Persistent Regression in the Authentication Service

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.