parallel-debugging

Parallel Debugging applies the Analysis of Competing Hypotheses methodology to organize multi-agent investigation of complex issues. It structures hypothesis generation across six failure mode categories, establishes evidence collection standards with citation requirements, and provides arbitration protocols to identify root causes while avoiding confirmation bias.

Parallel Debugging helps you systematically investigate complex bugs with multiple potential causes using competing hypotheses methodology.

AI-generated summary based on this skill's SKILL.md

★ 38,308 ⁮ 4,097 MIT updated 2026-07-22 by wshobson

Install

wshobson/agents/parallel-debugging · repository language: Python

Manual (git clone) CLI (skillfed)

git clone https://github.com/wshobson/agents
cp -r agents/plugins/agent-teams/skills/parallel-debugging ~/.claude/skills/parallel-debugging

npx skillfed install wshobson/agents/parallel-debugging

Frequently asked questions

AI-generated answers based on this skill's SKILL.md and metadata

How do I debug complex bugs that span multiple modules?

Parallel Debugging organizes investigation across multiple components by structuring hypothesis generation across six failure mode categories and establishing parallel investigation workflows. This approach lets you systematically explore different potential causes simultaneously rather than pursuing a single theory, reducing the risk of missing the actual root cause when bugs involve interactions between modules.

What is the competing hypotheses methodology for bug investigation?

Parallel Debugging applies the Analysis of Competing Hypotheses (ACH) methodology to structure root cause analysis. Instead of confirming a single theory, ACH requires you to generate multiple plausible explanations, collect evidence against each one, and use structured arbitration to identify which hypothesis is most consistent with your findings. This prevents confirmation bias by forcing systematic evaluation of alternatives.

How does parallel-debugging help avoid confirmation bias when debugging?

Parallel Debugging enforces evidence collection standards with citation requirements and competing hypotheses analysis. By requiring you to actively seek disconfirming evidence and evaluate multiple root causes in parallel rather than fixating on one theory, the framework prevents the natural tendency to interpret findings as support for your initial assumption. Confidence scoring helps arbitrate between plausible causes objectively.

What is a systematic debugging framework for hard-to-find bugs?

Parallel Debugging provides a structured root cause analysis framework that combines hypothesis generation across failure mode categories, evidence-based investigation with citation standards, and competing hypotheses methodology. This systematic approach organizes complex debugging workflows, establishes clear protocols for arbitrating between multiple plausible causes, and ensures you collect sufficient evidence before concluding what actually caused the issue.

How do I find the root cause of complex bugs with multiple potential causes?

Parallel Debugging structures root cause analysis by generating competing hypotheses across six failure mode categories, collecting evidence with citation requirements, and applying arbitration protocols with confidence scoring. This evidence-based methodology helps you systematically evaluate each potential cause, identify which explanations are most consistent with your findings, and avoid settling on an incorrect root cause prematurely.

What debugging methodology uses competing hypotheses and parallel investigation?

Parallel Debugging applies the Analysis of Competing Hypotheses (ACH) method to organize multi-agent investigation of complex issues. It structures hypothesis generation, establishes evidence collection standards, and provides arbitration protocols to identify root causes while avoiding confirmation bias. The framework is designed specifically for bugs with multiple potential causes across different components or modules.

SKILL.md

rendered from the published skill — quoted content, verbatim