investigate

Investigate failures through a four-phase workflow with evidence confirmation before fixes.

1|Updated Mar 27, 2026
One-click install
npx skills add https://github.com/yibeichan/schist --skill investigate-yibeichan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: investigate
Source: https://github.com/yibeichan/schist/tree/main/.claude/skills/investigate
Command: npx skills add https://github.com/yibeichan/schist --skill investigate-yibeichan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you debug failures systematically by locating the underlying root cause before making any changes, so you stop whack-a-mole fixes that break again later.

Core Features & Use Cases

  • Root-cause first debugging: Turns symptoms (errors, logs, and behavior) into a specific, testable hypothesis about what’s actually wrong.
  • Evidence-driven verification: Confirms or rejects hypotheses by adding targeted checks and re-running the reproduction path.
  • Controlled implementation: Applies a minimal change once the root cause is confirmed, then verifies it and documents the outcome.
  • Uses when you want to debug this/fix this bug/figure out why it’s broken, including troubleshooting regressions, intermittent issues, or configuration drift.

Quick Start

Use investigate when you paste the error and reproduce steps, then ask: Investigate this failure and tell me the most likely root cause and the smallest verified fix.

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 bug instead of just fixing symptoms?

Root cause investigation isolates the true source of a bug from its symptoms by forming testable hypotheses from errors and logs. It requires evidence confirmation before applying any fix, preventing recurring whack-a-mole patches.

What's the best way to debug an intermittent system failure or regression?

Debugging intermittent failures and regressions requires deterministic reproduction to confirm or reject hypotheses. By applying targeted checks and re-running the reproduction path, you verify the exact failure source before changing code.

How do I verify a fix without introducing new regressions?

Fix verification uses controlled implementation by applying a minimal change once the root cause is confirmed. It then re-runs the reproduction path with a regression-verification mindset to ensure behavior is fixed without new issues.

Does error triage work for troubleshooting configuration drift and unexpected errors?

Error triage handles troubleshooting configuration drift and unexpected errors by systematically isolating the underlying root cause. It transforms broken behavior into a specific, testable hypothesis about what is actually wrong.

What are the limits of hypothesis testing when debugging broken behavior?

Hypothesis testing for broken behavior requires deterministic reproduction and evidence confirmation before implementing fixes. Without controlled scope and a regression-verification mindset, changes may lack the evidence needed to truly resolve the issue.