mcpField
R
reliability-incident-learning-review

reliability-incident-learning-review

github.com/sylphxaiCloud & DevOps140 installsUpdated today
Repository

About reliability-incident-learning-review

A skill for teams reviewing reliability incidents and learning from their causes and effects. It structures impact, contributing factors, detection gaps, and follow-up work into an owned review with recurrence checks.

Key features

  1. 01Classifies incident types
  2. 02Documents timelines and customer impact
  3. 03Examines contributing factors and detection gaps
  4. 04Assigns actions, owners, and due dates
  5. 05Sets recurrence review cadence

Use cases

  1. 01Preparing a service outage postmortem
  2. 02Reviewing a release regression or dependency failure
  3. 03Tracking corrective actions after a customer-impacting incident

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sylphxai/skills --skill reliability-incident-learning-review

After installing, type/reliability-incident-learning-reviewin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlreferences/reliability-incident-learning-patterns.md

Details

From the project README

Reliability Incident Learning Review

Use this skill to convert a incident learning, postmortems, contributing factors, action items, recurrence prevention, and customer trust question into a concrete artifact with owners, gates, metrics, and recovery paths.

Workflow

  1. Identify incident type, customer impact, timeline, detection path, responders, affected promises, and unresolved risks.
  2. Read references/reliability-incident-learning-patterns.md.
  3. Classify incident as outage, degradation, data correctness, billing/auth, security/privacy, support process, release regression, or dependency failure.
  4. Define timeline, impact, contributing factors, detection gaps, comms review, action owners, due dates, recurrence guardrails, and learning distribution.
  5. Produce incident learning report, state machine, decision table, event schema, action checklist, and recurrence-review cadence.

When not to use

  • Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails

  • Do not reduce postmortems to one root cause when systems, process, monitoring, and decision factors interacted.
  • Do not close incident actions without evidence that recurrence risk changed.
  • Do not hide customer impact, missed detection, or communication gaps behind internal-only language.
  • Do not blame individuals instead of improving conditions, tools, and decision paths.

Output format

Incident context:
Incident type / customer impact / promise affected:

Incident learning plan:
| Area | Decision | Evidence | Risk | Owner |
| --- | --- | --- | --- | --- |

Factors, actions, and evidence:
- <item> -> <policy, metric, edge case, support note>

Recurrence and follow-through policy:
- <trigger> -> <action, communication, owner>

Comments

Markdown supported