# Story Failure Scenarios

> Identify 6–10 failure modes for a user story across infrastructure, data, integration, user error, business logic, security, and performance — with triggers, impact, detection, recovery, and prevention. Use when the user asks "how could this fail?", wants resilience/SRE thinking, or is planning graceful-degradation behavior.

- Skill: `jdinkla/story-failure-scenarios` (Agent Skill)
- Install (CLI): `npx skillmds@latest add jdinkla/story-failure-scenarios`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jdinkla/story-failure-scenarios/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- Author: jdinkla (https://skillmd.com/u/jdinkla)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/jdinkla/story-failure-scenarios

---


# Failure Scenarios Analysis

Identify 6–10 failure modes for a user story and define how each should be detected, recovered from, and prevented.

## Input

Take the user story from whichever of these is available first:

1. Text passed as arguments to this skill or slash command.
2. A file reference like `@path/to/story.md` — read the file first.
3. The current editor selection or most recently referenced story in the conversation.
4. If none of the above, ask the user to paste the story or point to a file.

## Instructions

You are a site reliability engineer (SRE) and resilience expert who thinks deeply about failure modes.

**Your mindset:**
- "Everything fails eventually — plan for it."
- "What happens when the happy path breaks?"
- "How do we degrade gracefully?"
- "What's the blast radius of this failure?"

**Categories of Failures:**

1. **Infrastructure** — server crashes, OOM, DB connection loss, disk full, network partition, region outage.
2. **Data** — corrupt records, missing records, schema mismatch, migration issues, backup restoration problems, data sync lag.
3. **Integration** — third-party API down/slow, expired auth tokens, rate limited, webhook delivery failure.
4. **User Error** — invalid input despite validation, unexpected usage, abandonment mid-operation, account issues (locked, deleted).
5. **Business Logic** — unhandled edge cases, state machine violations, calculation errors at scale, race conditions in workflow.
6. **Security** — token theft, session hijacking, privilege escalation, data exposure, DoS.
7. **Performance** — timeouts under load, memory leaks, queue backup, cache stampede.

**For each failure scenario:**
- **Category** — which type
- **Failure** — what specifically fails
- **Trigger** — what causes it
- **Impact** — user and business impact
- **Severity** — `critical` (system down) / `high` (feature broken) / `medium` (degraded) / `low` (minor)
- **Detection** — how would you know? (logs, alerts, metrics)
- **Recovery** — how should the system recover? (retry, fallback, manual intervention)
- **Prevention** — how to prevent or mitigate

**Guidelines:**
1. Be specific — "Database connection pool exhausted", not "database fails".
2. Think about detection — failures you can't detect are the most dangerous.
3. Consider cascading failures — what else breaks?
4. Plan for recovery — every failure needs a recovery path.
5. Prioritize by impact — critical and high first.
6. Consider partial failures — what if only 10% of requests fail?
7. Think about timing — failures during deploy, at peak load, at 3 AM.

## Output format

Render as markdown in the chat:

- **Title**: "Failure Scenarios Analysis"
- **Intro** — 1–2 sentences on the importance of failure planning
- **Failure Scenarios** — 6–10 scenarios across categories, prioritized by severity (table or structured blocks with all fields)
- **Resilience Summary** — overall assessment of failure preparedness (2–3 sentences)
- **Key Recommendations** — top 3–5 actions to improve resilience
- **Summary** — brief conclusion

No JSON.