diagnosing-bugs

Enforce a structured debugging workflow for hard-to-reproduce software bugs.

1|Updated Apr 14, 2026

Search Tags:#debugging #hypothesis testing #software testing #bug diagnosis #regression test #performance regression #reproducer

Ready to run

npx skills add https://github.com/hschne/pi-stuff --skill diagnosing-bugs-hschne

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: diagnosing-bugs
Source: https://github.com/hschne/pi-stuff/tree/main/skills/diagnosing-bugs
Command: npx skills add https://github.com/hschne/pi-stuff --skill diagnosing-bugs-hschne

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill eliminates the guesswork and wasted time when tackling hard-to-reproduce software bugs and performance regressions that lack clear error messages or consistent reproduction steps.

Core Features & Use Cases

Structured 6-phase debugging workflow: Enforces building a tight, deterministic feedback loop before forming hypotheses, preventing wasted effort on unproven theories.
Specialized guidance for edge cases: Includes tactics for non-deterministic flaky bugs, performance regressions, and scenarios where no automated repro can be built.
Use case example: If you have an intermittent 500 error that only occurs under high load, this skill walks you through building a load test repro, generating falsifiable hypotheses, and adding a regression test to lock in the fix.

Quick Start

Use the diagnosing-bugs skill to investigate the intermittent checkout failure that occurs when users apply a discount code.

Frequently Asked Questions about diagnosing-bugs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I debug an intermittent production error that only happens under high load?▼

Debugging intermittent production errors requires building a load test reproducer first. A structured workflow enforces deterministic reproduction, generates falsifiable hypotheses, and adds regression tests to lock in the fix.

What's the best way to diagnose flaky test failures without clear error traces?▼

Diagnosing flaky test failures involves specialized tactics for non-deterministic bugs. You enforce building a tight feedback loop before forming hypotheses, preventing wasted effort on unproven theories.

How do I find the root cause of a slow API endpoint performance regression?▼

Finding a slow API endpoint performance regression requires targeted instrumentation and falsifiable hypothesis testing. A structured 6-phase debugging workflow systematically isolates the bottleneck.

Can I diagnose software bugs when I cannot build an automated reproducer?▼

You can diagnose software bugs without an automated reproducer using specialized guidance for edge cases. The workflow adapts to scenarios where deterministic reproduction is impossible.

How do I stop hard-to-reproduce bugs from recurring after fixing them?▼

Stop hard-to-reproduce bugs from recurring by implementing regression tests. The debugging workflow requires implementing a regression test to verify the fix and prevent future recurrence.