Guide · Agent skill
How to use the Engineering Runbook agent skill (eng-runbook)
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Engineering Runbook skill (eng-runbook) teaches an AI agent to produce a single-page runbook for a service, the kind of quick reference an on-call engineer opens during an incident. This guide explains what agent skills are, how to install this one, how to ask for a runbook in plain language, and what to check before trusting the output. By the end you will be able to generate a consistent runbook layout and know where you must supply your own live information.
Engineering Runbooknexu-io/html-anything · Apache-2.0 Download the skill Quick start
- 1Download the ZIP. Download the eng-runbook ZIP from this page.
- 2Unzip into a skills folder. Unzip into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.
- 4Gather your real facts. Collect alert thresholds, dashboard links, commands and the on-call rotation for the service.
- 5Ask for a runbook. Request a runbook for the service in plain words, attaching or pasting your information.
- 6Review the output. Check every threshold, link, name and command against your live systems before relying on it.
What is an agent skill and what does this one make the agent do?
An agent skill is a folder of instructions that an AI agent loads when a task matches it. This skill is a template and layout guide: when asked, the agent writes one HTML-style page laid out as an engineering runbook.
The page follows a fixed structure: a service overview with topology and dependencies, an alerts table with severity, threshold and runbook link columns, dashboard link cards, common procedures in copyable monospace code blocks, the on-call rotation for this week and next, and an incident response checklist.
Install the eng-runbook skill
Download the ZIP for this skill from this page. You can install it in one of two ways.
Option 1, personal install: unzip the file into ~/.claude/skills/ so the skill is available across your projects.
Option 2, project install: unzip the file into the .claude/skills/ folder inside a specific project, so only that project uses it.
Option 3, Claude apps: upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- The package contains two files: SKILL.md (the instructions) and example.html (a sample page).
- Open SKILL.md after installing to confirm the skill is present and read what it covers.
How to trigger the skill with natural requests
You do not need a special command. Ask for a runbook in ordinary words that mention the runbook, the service and the kind of page you want. The skill is matched by its description, which covers the service overview, alerts table, dashboards, operational commands, on-call and incident checklist.
Give the agent the facts it cannot know. The skill does not pull live data, so include real alert thresholds, dashboard URLs, command details and rotation names in your request or in files the agent can read.
- Example: "Write an engineering runbook for our checkout service. Use the attached alert list and dashboard links."
- Example: "Make an on-call runbook page for the payments API with an alerts table and an incident response checklist."
- Example: "Turn this list of commands and this rotation schedule into a single-page SRE runbook."
What the skill needs and what it does not do
The skill lists no required tools or dependencies. It does not run scripts and does not use the network. It only shapes a document from the information you provide.
It will not connect to your monitoring, paging or infrastructure systems, and it will not run commands or fix anything automatically. Treat every command in the output as text to review before running.
- Requires: nothing listed
- Runs scripts: no
- Uses network: no
Safety checks before using a runbook in an incident
A runbook generated from guesses can send an engineer to the wrong fix. Because the skill does not verify facts, check every value against your live systems.
Review thresholds, severities, dashboard links, on-call names and every command against your real environment. Invented values are dangerous during an incident, so replace any placeholder or missing detail rather than trusting the agent's filling-in.
- Confirm each alert threshold and severity matches your alerting configuration.
- Click every dashboard and runbook link to confirm it opens the correct page.
- Verify on-call names and dates with your current rotation tool.
- Test copyable commands in a safe environment first where possible.
Tips for better runbooks
Give the agent one service per page. The layout works best when the overview, alerts and procedures all describe the same system.
Use example.html from the package as a reference for the expected look and structure, and ask the agent to match it.
- Provide dependencies explicitly so the topology section is accurate.
- Keep the incident checklist consistent across services so responders always know the next step.
What people use it for
- Drafting a one-page quick reference for a single production service.
- Standardising the layout of runbooks across several on-call teams.
- Documenting the alerts a service can raise, with severity and a link to each runbook.
- Collecting copyable operational commands in one page for faster incident response.
- Producing an incident response checklist to hand to a new on-call engineer.
Frequently asked questions
Is the eng-runbook skill free?
The skill is published under the Apache-2.0 licence, as listed for this item. Check the licence file in the source repository for the full terms before reusing it in your own products.
What do I need to install eng-runbook?
The skill lists no required dependencies. It runs no scripts and uses no network, so you only need to unzip the files into a skills folder or upload the ZIP in the Claude apps.
Where do I put the skill files?
Unzip them into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. You can also upload the ZIP under Settings → Capabilities → Skills in the Claude apps.
Will the skill pull my real alerts and dashboards?
No. The skill does not connect to your systems or fetch live thresholds, dashboards or rotation schedules. You must give the agent that information yourself.
Can the skill fix incidents automatically?
No. It only produces a document. It does not run commands on your infrastructure, so every command must be reviewed and run by a person.
How is this different from a runbook template or a monitoring tool?
A template is a fixed document layout, and a monitoring tool tracks live systems. This skill generates a runbook page from the information you supply, but it does not monitor anything or replace your alerting and paging tools.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use eng-runbook if you want an AI agent to produce a consistent, single-page on-call runbook layout quickly, and you can supply accurate facts yourself. Pick something else if you need the agent to read live alerts, dashboards or rotation schedules, or to run remediation automatically.
Download the skill