diagnosing-bugs

Diagnose software bugs through a six-phase feedback-loop methodology.

Updated Apr 8, 2026
One-click install
npx skills add https://github.com/tonomb/personal-pnl --skill diagnosing-bugs-tonomb
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagnosing-bugs
Source: https://github.com/tonomb/personal-pnl/tree/main/.agents/skills/diagnosing-bugs
Command: npx skills add https://github.com/tonomb/personal-pnl --skill diagnosing-bugs-tonomb

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill solves the problem of aimless debugging by enforcing a structured, evidence-based feedback loop that prevents premature hypothesis generation and ensures bugs are reproduced and verified before a fix is applied.

Core Features & Use Cases

  • Feedback Loop Construction: Provides a systematic framework to build tight, deterministic, and red-capable reproduction loops.
  • Structured Debugging: Guides the user through six distinct phases: building a loop, minimizing the repro, hypothesis generation, instrumentation, fixing, and post-mortem.
  • Use Case: When a user reports a complex performance regression or a non-deterministic crash, this skill forces the creation of a minimal, fast, and reliable reproduction command before any code changes are attempted.

Quick Start

Invoke the diagnosing-bugs skill to begin a structured debugging session for the reported issue by following the phase-based instructions provided in the documentation.

Frequently Asked Questions about diagnosing-bugs

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

FAQPage Schema
What is the best way to fix non-deterministic crashes and performance regressions?

Fixing non-deterministic crashes requires a structured, evidence-based feedback loop that mandates a deterministic, minimal reproduction case before attempting any code changes to prevent speculative fixes.

How do I build a deterministic reproduction loop for software debugging?

You can build a deterministic reproduction loop for software debugging by following a structured six-phase protocol that prioritizes evidence-based instrumentation, ensuring the reproduction command is minimal, fast, and reliably red-capable before fixing.

Why does premature hypothesis generation fail during root cause analysis?

Premature hypothesis generation fails during root cause analysis because it leads to aimless debugging and speculative code changes instead of relying on verified evidence gathered from a tight, deterministic reproduction loop.

What are the phases of structured troubleshooting for complex bugs?

Structured troubleshooting for complex bugs involves six distinct phases: building a reproduction loop, minimizing the repro, generating hypotheses, instrumenting for evidence, applying the fix, and conducting a post-mortem.

Can I use this debugging methodology for performance regressions across any platform?

You can use this debugging methodology for performance regressions across any platform, as it operates across the entire development lifecycle and enforces a mandatory feedback-loop-first approach independent of specific dependencies.

When should I not use speculative code changes for troubleshooting?

You should avoid speculative code changes for troubleshooting when dealing with complex bugs or performance regressions, as this structured approach requires verified instrumentation and a minimal reproduction case before any fix is applied.