diagnose

Diagnose hard bugs through a six-phase workflow from reproduction to fix.

Updated Nov 16, 2022
One-click install
npx skills add https://github.com/kos3nz/dotfiles --skill diagnose-kos3nz
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/kos3nz/dotfiles/tree/main/.config/agents/skills/diagnose
Command: npx skills add https://github.com/kos3nz/dotfiles --skill diagnose-kos3nz

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill provides a disciplined, phase-driven loop for diagnosing hard bugs and performance regressions, guiding the user from reproduction to hypothesis, instrumentation, and fix.

Core Features & Use Cases

  • Structured six-phase workflow: Build a feedback loop, reproduce, hypothesise, instrument, fix, and regression-test.
  • Guidance for deterministic repros, testability, and instrumentation to accelerate root-cause analysis.
  • Use cases include when a user reports bugs, performance regressions, or flaky behavior across a project.

Quick Start

Start by reproducing the bug, then follow the six phases: build the loop, reproduce, hypothesise, instrument, fix with regression tests, and cleanup.

Frequently Asked Questions about diagnose

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

FAQPage Schema
What is a disciplined debugging loop for hard bugs?

A disciplined debugging loop systematically resolves hard bugs and performance regressions through a six-phase workflow: build a feedback loop, reproduce, hypothesize, instrument, fix with regression tests, and cleanup.

How do I reproduce flaky behavior or crashes across a codebase?

To reproduce flaky behavior or crashes, you first build a deterministic feedback loop, then follow a structured workflow to hypothesize and instrument the codebase until the inconsistent behavior is reliably triggered for root-cause analysis.

What's the best way to instrument code for root-cause analysis?

The best way to instrument code for root-cause analysis is to apply a phase-driven workflow that generates artifacts and prompts targeted instrumentation after forming a hypothesis, accelerating the identification of crashes and slowdowns.

Can I use this structured debugging workflow for performance regressions?

Yes, you can use this structured debugging workflow for performance regressions, as it specifically applies when users report slowdowns, crashes, or inconsistent behavior across a project, guiding you through risk-managed diagnosis.

How do I add regression tests when fixing a hard bug?

You add regression tests during the fix phase of the debugging loop, ensuring the hard bug is resolved and the specific crash or flaky behavior is prevented from recurring in the codebase before final cleanup.