# Five Whys Failure Recovery

> On pipeline failure, drill five whys to root cause and log permanent system fixes — not prompt whack-a-mole. Use after hallucinations, format errors, or constraint misses. Scope boundary: human kaizen line → `ooda-lean-loop`; routing audit → `skill-library-audit`.

- Skill: `catcorner22/five-whys-failure-recovery` (Agent Skill)
- Install (CLI): `npx skillmds@latest add catcorner22/five-whys-failure-recovery`
- Raw SKILL.md: https://api.skillmd.com/api/skills/catcorner22/five-whys-failure-recovery/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: CatCorner22 (https://skillmd.com/u/catcorner22)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/catcorner22/five-whys-failure-recovery

---

# Five Whys failure recovery

**#6** · **Domain:** Toyota Production System · **Category:** quality-control · **Difficulty:** 🟡 Medium

## Core principle

Every defect is a system failure. Fix the system permanently.

## AI problem addressed

Retry with tweaked prompt fixes symptoms not causes.

## Implementation

1. Log failure artifact
2. Ask Why ×5 until schema/process gap found
3. Implement structural fix (schema, gate, preflight)
4. Add to failure case library
Example: missing file → orchestration passes text not metadata → add inter-step schema.

## Boundaries

- Prototype as prompt scaffold (🟢) before full pipeline middleware (🟡/🔴)
- Category router: **`ai-transfer-quality-control`**
- Catalog: **`ai-transfer-ecosystem-primer`**
- Runtime plugin id: `root_cause_drill`