challenge

Identifies falsifiable claims in debugging hypotheses and seeks counter-evidence in code and artifacts.

7|15|Updated Jan 23, 2026

Search Tags:#debugging #root cause analysis #experiment design #hypothesis testing #adversarial review #codebase search #log investigation

Ready to run

npx skills add https://github.com/openshift-eng/edge-tooling --skill challenge-openshift-eng

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: challenge
Source: https://github.com/openshift-eng/edge-tooling/tree/main/plugins/challenge/skills/challenge
Command: npx skills add https://github.com/openshift-eng/edge-tooling --skill challenge-openshift-eng

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you rigorously challenge the most recent hypothesis or root-cause explanation in a debugging or investigation conversation, so weak or unverified claims don’t harden into incorrect conclusions.

Core Features & Use Cases

Adversarial review with no validation: attacks the hypothesis directly and avoids endorsing it while still producing actionable ways to verify or falsify it.
Counter-evidence investigation first: breaks the hypothesis into falsifiable claims, searches the codebase and available artifacts for contradictions, and records how each finding changes the hypothesis strength.
Experiment-driven resolution: proposes concrete commands and log/code checks that would confirm or deny each counter-argument.

Quick Start

Ask the skill to challenge the latest hypothesis by generating an adversarial review with counter-evidence searches and definitive experiments.

Frequently Asked Questions about challenge

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

FAQPage Schema

How do I rigorously challenge a root cause hypothesis during debugging?▼

To challenge a root cause hypothesis, you break it into falsifiable claims, search the codebase for counter-evidence, and output structured counter-arguments. This adversarial review avoids validating the hypothesis while actively seeking contradictions in logs and code.

What is adversarial review in root cause analysis?▼

Adversarial review in root cause analysis is a technique that attacks a debugging hypothesis directly by identifying falsifiable claims. It searches code and artifacts to find contradictions, ensuring weak explanations do not harden into incorrect conclusions.

Can I use hypothesis testing for CI failure triage and runtime incident investigation?▼

Yes, hypothesis testing applies to CI failure triage and runtime incident investigation. It evaluates multiple plausible causal theories by proposing concrete commands and log checks to confirm or deny each counter-argument found during the investigation.

How do I design experiments to confirm or deny a debugging hypothesis?▼

You design experiments by proposing concrete commands and log checks that verify or falsify each counter-argument. This experiment-driven resolution generates definitive tests to confirm or deny the challenged elements of the root-cause explanation.

What is the best way to investigate counter-evidence for a root cause explanation?▼

The best way to investigate counter-evidence is performing grep and read-based codebase searches for each claim within the hypothesis. This structured approach records how each finding changes the hypothesis strength without endorsing the original explanation.

When should I avoid validating a hypothesis during codebase search?▼

You should avoid validating a hypothesis during a codebase search when multiple causal theories are plausible. Instead, perform an adversarial review to actively seek contradictions and ensure you are not simply confirming your existing bias.