investigate

Identify root causes of software bugs through deterministic debugging and evidence collection.

Updated Mar 28, 2026
One-click install
npx skills add https://github.com/mattothomas/sp26_hackpsu --skill investigate-mattothomas
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: investigate
Source: https://github.com/mattothomas/sp26_hackpsu/tree/main/.claude/skills/gstack/investigate
Command: npx skills add https://github.com/mattothomas/sp26_hackpsu --skill investigate-mattothomas

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Root-cause debugging is a structured approach to diagnose why a bug happens instead of chasing symptoms, reducing fix cycles by focusing on the real cause.

Core Features & Use Cases

  • Four-phase methodology: investigate, analyze, hypothesize, implement.
  • Guided symptoms collection, deterministic reproduction, and evidence-backed root-cause hypotheses.
  • Applicable to errors, crashes, or unexpected behavior across codebases, from scripts to services.

Quick Start

Describe the symptoms, reproduce the issue deterministically, and follow the four phases to identify the root cause before attempting a fix.

Frequently Asked Questions about investigate

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

FAQPage Schema
What is root-cause debugging and how does it differ from fixing symptoms?

Root-cause debugging diagnoses why a bug happens rather than chasing symptoms, reducing fix cycles by focusing on the actual cause. It uses a structured approach to trace evidence back to the origin.

How do I systematically find the root cause of a software bug?

To find a root cause systematically, follow a four-phase methodology: collect symptoms, reproduce the issue deterministically, analyze patterns and test hypotheses, then implement the fix.

Can I use shell commands to reproduce bugs deterministically across codebases?

Yes, deterministic reproduction supports scope locking and traceable evidence using Bash commands and shell helpers. This requires access to Bash, Read, Write, Edit, Grep, Glob, and WebSearch tools.

What is the best way to investigate error messages and stack traces?

The best way to investigate error messages and traces is guided symptoms collection followed by code reading. This process locks scope and builds evidence-backed root-cause hypotheses before attempting fixes.

Does root-cause debugging work for crashes and unexpected behavior in scripts?

Root-cause debugging is applicable to errors, crashes, or unexpected behavior across codebases, from scripts to services. It guides pattern analysis and hypothesis testing for any software issue.