investigate

Guide root-cause debugging with evidence collection and hypothesis validation.

1|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/bhianerd/aimemory --skill investigate-bhianerd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: investigate
Source: https://github.com/bhianerd/aimemory/tree/main/.claude/skills/gstack/investigate
Command: npx skills add https://github.com/bhianerd/aimemory --skill investigate-bhianerd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps you debug issues by forcing a disciplined root-cause investigation, so you don’t apply fixes that only hide symptoms.

Core Features & Use Cases

  • Four-phase investigation workflow: investigate, analyze, hypothesize, and implement with explicit checkpoints.
  • Evidence-first hypothesis testing: validates suspected causes with targeted log/assertion/debug output before changing code.
  • Scope control to prevent scope creep: optionally restricts edits to the narrowest affected directory when a “freeze” boundary is available.
  • Regression-minded fixes: requires a regression test and full test-suite verification before declaring success.
  • Guided escalation: stops after repeated failed attempts and requests architectural review or logging/monitoring instead of guessing.

Quick Start

Tell the AI to investigate the bug or error you’re seeing, including the exact message and steps to reproduce.

Frequently Asked Questions about investigate

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

FAQPage Schema
How do I find the root cause of a software bug instead of just fixing the symptom?

Root cause debugging requires collecting evidence, reading relevant code paths, and verifying hypotheses through targeted instrumentation before implementing changes. This systematic approach prevents symptom-only fixes by enforcing staged investigation, hypothesis validation, and verified fixes with regression tests and full test-suite results.

What is the best way to debug regression scenarios that need deterministic reproduction?

Debugging regression scenarios requires deterministic reproduction and careful change control. A disciplined investigation workflow applies evidence-first hypothesis testing with targeted log output, debug-scope locking to prevent scope creep, and mandatory regression tests verified by full test-suite execution before declaring success.

How do I stop scope creep when investigating unexpected behavior in my codebase?

Scope creep during code investigation is controlled by optionally restricting edits to the narrowest affected directory using a freeze boundary. This scope-locking mechanism ensures changes remain focused on the verified root cause rather than expanding into unrelated code paths during the debugging workflow.

When should I escalate a debugging session for architectural review instead of guessing?

Escalation for architectural review or logging and monitoring requests should occur after repeated failed debugging attempts. Guided escalation stops the investigation process to prevent guessing, ensuring that persistent software failures are routed to architectural review rather than applying unverified code changes.

Does root cause debugging require regression testing before implementing a fix?

Regression testing is required before completing root cause debugging fixes. The investigation workflow mandates writing regression tests and verifying full test-suite results to confirm the fix resolves the software failure without introducing new unexpected behavior into the codebase.

Can I use instrumentation to validate hypotheses during code investigation?

Instrumentation is used to validate suspected causes during code investigation by adding targeted log, assertion, and debug output before changing code. This evidence-first hypothesis testing ensures the root cause is confirmed through deterministic reproduction rather than assumed before implementing fixes.