diagnosing-bugs

Diagnose hard bugs and performance regressions through a structured feedback loop.

Updated Jul 9, 2026
One-click install
npx skills add https://github.com/lyingparachute/skills --skill diagnosing-bugs-lyingparachute
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagnosing-bugs
Source: https://github.com/lyingparachute/skills/tree/main/diagnosing-bugs
Command: npx skills add https://github.com/lyingparachute/skills --skill diagnosing-bugs-lyingparachute

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill provides a structured approach to diagnose hard bugs and performance regressions, enabling users to quickly identify and fix issues.

Core Features & Use Cases

  • Diagnosis Loop: Offers a comprehensive loop for diagnosing bugs, including feedback loop construction, tightening the loop, and handling non-deterministic bugs.
  • Reproduce + Minimise: Assists in reproducing and minimizing the bug to its smallest scenario, making it easier to debug.
  • Hypothesise: Generates 3-5 ranked hypotheses for the cause of the bug, ensuring a systematic approach to debugging.
  • Instrument: Provides guidance on instrumenting the code for debugging, including using debuggers, targeted logs, and performance measurement.
  • Fix + Regression Test: Offers steps to fix the bug and create a regression test to prevent future occurrences.
  • Cleanup + Post-mortem: Ensures proper cleanup and documentation of the debugging process for future reference.

Quick Start

Run the 'diagnosing-bugs' skill to start diagnosing a hard bug in your codebase.

Frequently Asked Questions about diagnosing-bugs

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

FAQPage Schema
How do I diagnose hard bugs and performance regressions systematically?

To systematically diagnose hard bugs, construct a feedback loop, reproduce and minimize the bug scenario, generate ranked hypotheses, instrument the code, and create a regression test after fixing. This structured approach ensures quick identification and resolution.

What is the best way to debug non-deterministic bugs in my codebase?

Debugging non-deterministic bugs requires tightening the feedback loop and systematically reproducing the issue. By minimizing the bug scenario and using targeted code instrumentation, you can isolate the root cause and generate accurate regression tests.

How do I generate hypotheses for a code bug I cannot easily reproduce?

Generate 3-5 ranked hypotheses for the bug cause by analyzing the minimized reproduction scenario. Instrument the code with debuggers and targeted logs to verify each hypothesis systematically until the root cause is identified.

How do I fix a performance regression and prevent it from happening again?

Fix a performance regression by measuring performance during code instrumentation, applying the fix, and creating a regression test. Proper cleanup and post-mortem documentation ensure the debugging process is recorded for future reference.

Do I need a deep understanding of my codebase to use a systematic debugging approach?

Yes, a systematic debugging approach requires a clear understanding of the codebase and debugging techniques. This foundational knowledge is necessary to effectively construct feedback loops, generate accurate hypotheses, and instrument code for performance measurement.

What should I do after fixing a difficult bug to ensure proper code hygiene?

After fixing a difficult bug, perform cleanup and document the debugging process in a post-mortem. Creating a regression test is also essential to verify the fix and prevent future occurrences of the same issue.