Structured root cause research skill

Structured Root-Cause Research

Use this skill to turn a vague or complex problem into a grounded decision. Favor evidence, clear assumptions, confidence levels, and verification steps over long hidden reasoning.

Do not expose private chain-of-thought. Provide a concise reasoning summary through root-cause chains, evidence tables, flow maps, tradeoffs, and recommendations.

Workflow

1. Frame The Problem

State the decision this analysis should support:

• Observed symptom, question, or decision
• Expected behavior or target outcome
• Affected users, systems, or business goal
• Known constraints, deadlines, and risks
• Success criteria for a useful answer

If the task is small, keep this to one or two sentences.

2. Build An Evidence Ledger

Gather only evidence needed for the decision:

• Local evidence: code paths, configs, logs, tests, schemas, screenshots, traces
• External evidence: official docs, standards, reputable engineering writeups, current customer or market evidence
• Dates, versions, and environment details when facts may change
• Counterexamples or conflicting evidence

Prefer primary sources for correctness, security, compliance, or money-sensitive claims. Mark unsupported or single-source claims as low confidence.

3. Run Evidence-Based 5 Whys

Use 5 Whys as a compact reasoning summary, not as a private reasoning transcript:

Symptom: [what is happening]
Why 1: [immediate cause, tied to evidence]
Why 2: [deeper system cause, tied to evidence]
Why 3: [design/process/root cause, tied to evidence]
Root cause: [actionable cause]
Confidence: high | medium | low

Stop when the cause is specific enough to change code, process, design, positioning, or the next experiment. Do not force exactly five levels if fewer or more are warranted.

4. Map Relevant Flows

Map only the flows needed to explain the issue:

Flow	Source	Control	Sink	Failure Mode
Data	User/API/input	Validation/transform	DB/output	Wrong state or leakage
Control	User action/job/event	Guard/policy	Side effect	Unexpected execution
Error	Exception/timeout	Fallback/retry	Log/UI/user	Silent failure or bad recovery
Trust	Actor/content/config	Permission boundary	Privileged action	Boundary bypass

For security work, identify source, control, sink, impact, and authorization scope. For customer or product research, replace sink with customer outcome.

5. Compare Against Better Practice

Use comparison only when it changes the decision:

Aspect	Current	Better Practice	Gap	Evidence
Design	...	...	...	file/source
Reliability	...	...	...	test/log/source
Security	...	...	...	policy/CWE/OWASP/source
Maintainability	...	...	...	code pattern/source

Avoid vague "SOTA" claims. Name the standard, source, product pattern, or engineering practice being used for comparison.

6. Recommend The Smallest Durable Change

Recommend a solution or decision that addresses the root cause:

• State assumptions and tradeoffs
• Name what should not change
• Prefer existing project patterns and low-blast-radius changes
• Add verification proportional to risk
• Include rollback or follow-up when impact is high

Output Formats

Quick Analysis

## Problem
[One-line problem statement]

## Root Cause
[2-4 sentence reasoning summary with confidence]

## Recommendation
[Specific action and verification]

Deep Analysis

## Problem
[Symptom, expected outcome, scope]

## Evidence
| Evidence | Source | Confidence |
| --- | --- | --- |
| ... | ... | high/medium/low |

## Root Cause Summary
Why 1: ...
Why 2: ...
Why 3: ...
Root cause: ...
Confidence: ...

## Flow Map
[Table or concise diagram]

## Comparison
| Aspect | Current | Better Practice | Gap |
| --- | --- | --- | --- |
| ... | ... | ... | ... |

## Recommendation
[Minimal durable solution]

## Verification Plan
- [ ] ...
- [ ] ...

Quality Bar

• Do not jump from symptom to fix without evidence.
• Do not present private chain-of-thought; summarize reasoning.
• Do not use 5 Whys as a single-cause story when multiple causes are plausible.
• Do not use best-practice language without naming the source or pattern.
• Do not recommend broad rewrites when a smaller durable change works.
• Do not over-research low-risk tasks.
• Do not fabricate sources, benchmarks, customer demand, or security impact.
• Separate facts, inferences, and opinions.
• Preserve uncertainty when evidence is incomplete.

Example Uses

Backend Bug

Problem: Hibernate fails on startup with "Not a managed type".
Root cause summary: A Spring Data JPA repository points at a domain model instead of a JPA entity, violating the clean architecture persistence boundary.
Recommendation: Change the JPA repository generic to the infrastructure entity, keep the domain repository port unchanged, and add or adjust a repository adapter test.

Security Review

Problem: A project config appears to trigger privileged behavior during a headless CLI run.
Evidence: Policy docs, local reproduction, negative control, fake canary only.
Root cause summary: Source-controlled config may cross a trust boundary before the expected control is applied.
Recommendation: Report only if the behavior contradicts documented restrictions and has a concrete authorized source-control-to-privileged-sink path.

Customer/Offer Research

Problem: A technical service offer is too broad to sell.
Root cause summary: The buyer cannot map "AI/security/full-stack" to a concrete pain, deliverable, price, or risk reduction.
Recommendation: Package one narrow offer with a sample report, lead scoring rubric, and a 10-lead outreach test.