Generate operational runbooks for incidents and deployments

Runbook Generator creates operational procedures that operators can follow during incidents, deployments, and maintenance. It clear prerequisites, ordered actions, exact commands, expected outputs, verification steps, timestamps and version information, permissions, dependencies, and explicit rollback instructions. Procedures should be tested in a non-production environment before use and should include safety checks and confirmation prompts for risky actions.

Guidance covers specialized incident-response, deployment, and maintenance runbooks, as well as executable shell procedures. An incident example addresses database connection-pool exhaustion with immediate actions and escalation triggers. Deployment content separates pre-deployment preparation from execution; maintenance content includes system preparation and automation. A health-check script demonstrates how a procedure can combine service status and connectivity checks.

Use this skill when turning operational knowledge into repeatable steps for an on-call responder or system administrator. It is useful when vague instructions, hidden assumptions, or missing rollback paths could make recovery slower or less safe. It does not validate a command against a live environment automatically; commands and recovery steps need review and safe testing in the target context.

Inputs include the procedure type, system and version, prerequisites, required access, dependencies, operational sequence, expected observations, escalation conditions, and recovery path. Outputs can include a human-readable checklist, an executable script, or both. Each step should name the action and what success looks like. Rollback procedures should be verified, and contact information and diagrams should be maintained as systems change.

the quality checklist includes:

All commands tested in safe environment
Rollback procedures verified

The maintenance guidance recommends quarterly review, updates after system changes, validation during disaster-recovery exercises, feedback from operators, and version control. Avoid vague verbs such as “restart the system” without a specific target and check. Include alternatives where a single point of failure exists, and avoid commands without expected outputs or prerequisites.

This skill is for SREs, operations teams, system administrators, and service owners who need reliable procedures under time pressure. It helps make runbooks actionable, reviewable, and maintainable, while connecting response steps to verification and escalation rather than leaving operators to infer what happens next.

Use the guidance when its stated workflow matches the problem. It provides an organizing method and examples rather than a substitute for local requirements, tested configuration, or domain-specific review. Adapt choices to the real system and verify the resulting procedure or pipeline before relying on it.

Start with the relevant system, its constraints, the desired outcome, and the behaviors that need to be represented. The guidance turns those inputs into a structured plan, configuration, or workflow, with checks that make outcomes observable. Use the examples as patterns and keep only the components supported by the actual environment.

Examples include this example excerpt:

#!/bin/bash
# Health Check Runbook Script
# Step 1: Check service status
# Step 2: Verify connectivity

Examples and named tools are described above; select the ones that match the implementation rather than combining unrelated options.

This is for practitioners responsible for the corresponding engineering or operational work. It helps make choices explicit, preserve repeatability, and communicate a process that can be reviewed and maintained.