Runbook generator

Write incident runbooks with symptom, verification, fix steps, and escalation paths.

49|11|Updated Jul 31, 2026
One-click install
npx skills add https://github.com/vstorm-co/agenticos --skill runbook-generator-vstorm-co
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Runbook generator
Source: https://github.com/vstorm-co/agenticos/tree/main/backend/app/core/catalog/skill_gallery/software/runbook-generator
Command: npx skills add https://github.com/vstorm-co/agenticos --skill runbook-generator-vstorm-co

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? On-call engineers woken at 3am often face runbooks that assume context they do not have, contain vague steps like "investigate the logs", or hide rollback instructions in other documents. This Skill produces runbooks written for the least-context person on the rota, so any responder can act immediately without prior knowledge. ## Core Features & Use Cases - Standardized Structure: Enforces a six-part shape covering Symptom, Verify, Impact, Fix, If-that-fails, and Afterwards. - Copy-Pasteable Commands: Requires exact commands with <angle bracket> placeholders, expected output after each step, and reversal instructions before every destructive action. - Named Escalation: Mandates naming a specific rota or person instead of escalating to an anonymous team. - Use Case: After a recurring database failover alert, generate a runbook that lets a junior engineer confirm the symptom with one command, execute numbered fix steps, and know exactly who to wake if the fix fails. ## Quick Start Write a runbook for the "disk usage above 90%" alert on our production database server, including verification, fix steps, and rollback.

Frequently Asked Questions about Runbook generator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a runbook for on-call incident response?▼

Structure the runbook in six sections: Symptom in the alert's own words, one Verify command, Impact for severity, numbered Fix steps with exact commands, an If-that-fails branch, and Afterwards checks. Write for the least-context person on the rota.

What should a good incident runbook include?▼

A runbook should include the symptom as the alert phrases it, a single verification command, impact scope, copy-pasteable fix commands with expected output, rollback instructions before destructive steps, and a named escalation rota.

How do I handle rollback steps in a runbook?▼

Place the reversal instructions immediately before every destructive command, stating what it destroys and how to undo it. Never leave rollback procedures in a separate document, since responders at 3am will not find them.

Why do runbooks fail during real incidents?▼

Runbooks fail when they contain vague steps like "investigate the logs", assume tools are installed, or escalate to an anonymous team instead of a named rota. Each step must say how to tell it worked.

When should I not use a runbook for incident response?▼

Runbooks fit known, recurring failure modes with deterministic fixes. For novel incidents with unknown root cause, a runbook cannot replace live diagnosis, so it should instead name who to wake for that class of problem.