# Failure Mode Scanner

> Identify the most likely ways a plan, change, rollout, or system could fail before execution. Use when the next move carries meaningful downside and the main need is to surface realistic failure modes rather than generic caution.

- Skill: `alubiama/failure-mode-scanner` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add alubiama/failure-mode-scanner`
- Raw SKILL.md: https://api.skillmd.com/api/skills/alubiama/failure-mode-scanner/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Alubiama (https://skillmd.com/u/alubiama)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/alubiama/failure-mode-scanner

---


# Failure Mode Scanner

## Goal

Surface the most likely and most damaging failure modes before action.

This skill does not replace full incident planning. It identifies where the plan can break and what deserves a guardrail, rollback, or monitor.

## When To Use

- before risky implementation or rollout
- when hidden downside may be larger than visible upside
- when safety planning needs concrete failure paths

## Scope Boundaries

In scope:

- identify likely failure modes
- rank by severity and plausibility
- connect failure modes to mitigations

Out of scope by default:

- exhaustive threat models
- full remediation runbooks

## Workflow

1. State the intended change or action.
2. Identify key assumptions and dependencies.
3. Name the likely failure modes.
4. Rank by probability and impact.
5. Map each serious mode to a mitigation, guardrail, or rollback need.

## Output Contract

1. `planned_action`
2. `failure_modes`
3. `highest_risk_modes`
4. `severity_logic`
5. `mitigations_needed`
6. `recommended_followup`
7. `next_step`

## Guardrails

- Do not fill the list with low-value hypotheticals.
- Prefer realistic operational failures over dramatic edge cases.
- Keep the scan decision-relevant.