skills.rest
cogine-ai
cogine-aiOFFICIAL
Explore all skills by this author →
WebsiteGitHub

investigate

Diagnose software failures through a four-phase root-cause analysis methodology.

3|1|Updated Mar 17, 2026
Search Tags:#debugging#investigate#root-cause#reproduction#pattern-analysis#flaky-tests
One-click install
npx skills add https://github.com/cogine-ai/cogine-dev-skillset --skill investigate-cogine-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: investigate
Source: https://github.com/cogine-ai/cogine-dev-skillset/tree/main/skills/investigate
Command: npx skills add https://github.com/cogine-ai/cogine-dev-skillset --skill investigate-cogine-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Systematic, evidence-based debugging workflow that identifies and fixes root causes behind bugs, failures, and flaky tests using a disciplined, four-phase approach.

Core Features & Use Cases

  • Phase 1: Root Cause Investigation
  • Collect Symptoms: Gather evidence before forming hypotheses, including error messages and reproduction conditions
  • Reproduce: Write down exact steps to trigger the failure
  • Narrow the Scope: Use binary-search style analysis to identify the module, data, or environment causing the issue
  • Phase 2: Pattern Analysis
  • Match common bug patterns (race conditions, nil propagation, state corruption, integration failure, config drift, stale caches)
  • Phase 3: Hypothesis Testing
  • For each hypothesis, predict, test, observe, and conclude with minimal changes and controlled experiments
  • Phase 4: Implementation
  • Fix root cause with minimal diff and regression tests

Quick Start

Describe symptoms, reproduce the failure, and apply the four-phase root-cause analysis to confirm the root cause.

Frequently Asked Questions about investigate

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

FAQPage Schema
How do I perform structured root-cause analysis for software failures?▼

Structured root-cause analysis for software failures follows a four-phase methodology: investigating symptoms, analyzing patterns, testing hypotheses, and implementing fixes with reproducibility checks to ensure evidence-based conclusions.

What is the best way to debug flaky tests and unexpected behavior in codebases?▼

Debugging flaky tests and unexpected behavior requires gathering evidence, reproducing exact failure conditions, narrowing scope via binary-search, and matching common bug patterns like race conditions or state corruption to identify the root cause.

How do I systematically reproduce bugs and narrow down the causing module?▼

To systematically reproduce bugs, write down exact steps to trigger the failure, then use binary-search style analysis to isolate the specific module, data, or environment causing the issue before forming hypotheses.

What common bug patterns should I look for during pattern analysis?▼

Common bug patterns to look for during pattern analysis include race conditions, nil propagation, state corruption, integration failures, config drift, and stale caches that cause unexpected software behavior.

How do I test hypotheses and fix root causes without introducing regressions?▼

Test hypotheses by predicting, testing, and observing outcomes with minimal changes and controlled experiments, then fix the root cause with a minimal diff and add regression tests to prevent future failures.

Related Skills

Explore More Software Engineering Skills→
jk278jk278

debugging-assistant

Diagnose software bugs from stack traces and error logs.

Community
Advanced
a29paula29paul
1

gstack-openclaw-investigate

Enforce root cause investigation before bug fixes in software debugging workflows.

Community
Advanced
thejefflarsonthejefflarson
20

race-condition

Detect and prevent TOCTOU and race conditions in code paths.

Community
Intermediate
sachio222sachio222

systematic-debugging

Enforce root-cause investigation before implementing fixes for software issues.

Community
Advanced