Skip to content
Back to skills

Root Cause Analysis

ASecurity

Generate a canonical RCA artifact for an incident or failure.

  • 8 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 2, 2026
ai-agents

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned September 2, 2026

npx -y skills add Eliyce/paqad-ai --skill root-cause-analysis --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Root Cause Analysis?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Root Cause Analysis
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/eliyce-root-cause-analysis/badge)](https://www.skillsdirectory.com/skills/eliyce-root-cause-analysis)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: root-cause-analysis
description: Generate a canonical RCA artifact for an incident or failure.
model_tier: reasoning
triggers:
  - workflow:
      - root-cause-analysis
cacheable: false
cache_key_inputs: []
output_format: markdown
input_schema:
  request_text:
    type: string
    required: true
    description: Incident or problem statement to analyze.
  stack_snapshot_path:
    type: path
    required: false
    description: Optional stack snapshot to ground the investigation.
---

## What It Does

Produces a canonical root cause analysis document with a fixed section set so investigations stay comparable across stacks and incidents.

## Use This When

Use this when the workflow is explicitly root-cause-analysis and the team needs a durable artifact that captures the problem, evidence, root cause, and solution in one place.

## Inputs

- Read the incoming problem statement first.
- Read the current stack snapshot, logs, traces, tests, or incident notes when available.
- Use stack-specific tools and rules only as investigation inputs, not as a reason to vary the RCA structure.

## Procedure

1. Restate the problem clearly and identify the affected stack context.
2. Capture symptoms, impact, and timeline evidence before writing a conclusion.
3. Separate confirmed root causes from contributing factors and open questions.
4. Record the remediation in the `Solution` section even when it is proposed rather than already deployed.
5. End with verification steps and follow-up actions needed to prevent recurrence.

## Output Contract

- Match `assets/output.template.md` exactly: 8 canonical sections in the order from `assets/section-order.txt`.
- `## Solution` is always present, even when the remediation is proposed not deployed.
- Output must pass `scripts/lint-output.sh` (exit 0) — it enforces the section order strictly.

## Escalate / Stop Conditions

- Ask when there is not enough evidence to distinguish root cause from speculation.
- Warn when the available evidence conflicts across logs, tests, or runtime observations.
- Do not present a guess as a confirmed root cause.

## Resources

- `references/rca-sections.md`
- `scripts/lint-output.sh`
- `assets/output.template.md`
- `assets/section-order.txt`
- `agents/openai.yaml`

Files in this skill

  • SKILL.md2.2 KB
  • agents/openai.yaml130 B
  • assets/output.template.md709 B
  • assets/section-order.txt120 B
  • references/rca-sections.md218 B
  • scripts/lint-output.sh1.2 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.