diagnose

Diagnose bugs and performance regressions through structured reproduce, hypothesize, instrument, and verify phases.

5|Updated Oct 18, 2025
One-click install
npx skills add https://github.com/rexshihaoren/rexyDigest --skill diagnose-rexshihaoren
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagnose
Source: https://github.com/rexshihaoren/rexyDigest/tree/main/docs/ai/skills/diagnose
Command: npx skills add https://github.com/rexshihaoren/rexyDigest --skill diagnose-rexshihaoren

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill provides a disciplined workflow to diagnose hard bugs and performance regressions by creating deterministic feedback loops that yield reproducible signals for fixes.

Core Features & Use Cases

  • Structured phases: reproduce → hypothesise → instrument → fix → regression test, ensuring a clear, auditable path from issue to fix.
  • Reproducible loops: guidance on building repeatable test setups and signals to identify root causes quickly.
  • Guardrails & documentation: checkpoints, logs with unique prefixes, and explicit success criteria to keep debugging efforts science-based and auditable.

Quick Start

Open a reproducible loop to diagnose a bug quickly by reproducing, hypothesizing, instrumenting, and validating fixes.

Frequently Asked Questions about diagnose

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

FAQPage Schema
What is a systematic debugging loop for diagnosing hard bugs?

A systematic debugging loop diagnoses hard bugs through structured phases: reproduce, hypothesize, instrument, fix, and regression test. This disciplined workflow creates deterministic, reproducible signals that guide fixes and keep debugging efforts auditable and science-based.

How do I find the root cause of a performance regression using hypothesis testing?

To find the root cause of a performance regression, build a reproducible test setup, form a hypothesis, instrument the code for fail-fast signals, and verify the fix. Staged hypothesis testing ensures experiments remain deterministic and auditable throughout the diagnosis process.

What is the best way to reproduce bugs for root-cause analysis?

The best way to reproduce bugs for root-cause analysis is to build repeatable test setups that generate fail-fast signals. Using structured phases and logs tagged with unique prefixes keeps debugging experiments deterministic and provides clear, reproducible loops for identifying root causes quickly.

Can I use this debugging workflow for both unit tests and end-to-end scenarios?

Yes, you can use this debugging workflow for both unit tests and end-to-end scenarios. The structured phases of reproducing, hypothesizing, instrumenting, fixing, and regression testing apply across debugging tasks, using fail-fast signals and staged hypothesis testing to guide fixes in any context.

How do instrumentation and checkpoints keep bug diagnosis auditable?

Instrumentation and checkpoints keep bug diagnosis auditable by enforcing explicit success criteria and tagging logs with unique prefixes. This disciplined feedback loop ensures every experiment remains deterministic, traceable, and science-based from initial reproduction through final regression testing.

When should I not use a structured debugging loop for bug reproduction?

You should avoid a structured debugging loop when a bug is trivial or immediately obvious from a quick code review, as the overhead of formal reproduction, hypothesis testing, and instrumentation is designed for hard bugs and complex performance regressions requiring deterministic, auditable investigation.