Applies Google SRE principles including SLOs and error budgets to improve reliability.

Install

mkdir -p .claude/skills/sre-practices && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/14464" && unzip -o skill.zip -d .claude/skills/sre-practices && rm skill.zip

Installs to .claude/skills/sre-practices

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Site Reliability Engineering practices from Google - the company that invented SRE. Master SLOs, error budgets, incident response, and toil elimination. Use when designing reliable systems, implementing SRE practices, or improving operational excellence. Learn from the team that runs Google Search, Gmail, and YouTube at billions of users scale.
346 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Advanced

Key capabilities

  • Define and measure service quality with SLIs, SLOs, and SLAs.
  • Manage error budgets to balance reliability and velocity.
  • Automate manual, repetitive work to eliminate toil.
  • Monitor and alert on user-facing symptoms using golden signals.
  • Conduct blameless postmortems and reduce Mean Time To Recovery (MTTR).
  • Plan for growth, forecast demand, and optimize resource usage.

How it works

This skill applies Site Reliability Engineering principles from Google, focusing on defining SLOs, managing error budgets, eliminating toil, and improving incident response to achieve high service reliability.

Inputs & outputs

You give it
Service reliability requirements, operational data, or incident reports.
You get back
Defined SLOs, error budget reports, toil automation plans, incident response strategies, and capacity plans.

When to use sre-practices

  • Define service objectives
  • Implement error budgets
  • Improve incident response

About this skill

SRE Practices - Google Site Reliability Engineering

Expert: Alex Kim (Google SRE, 11 years) Level: 10/10 - Google invented SRE

Overview

Site Reliability Engineering from Google - what happens when you ask a software engineer to design an operations team. Not traditional ops or DevOps - applying software engineering to infrastructure.

Google runs services for billions (Search, Gmail, YouTube, Maps) with 99.99%+ uptime. These practices made that possible.

Core SRE Principles

1. Embrace Risk

100% uptime is the wrong target. Use error budgets to balance reliability vs velocity.

2. Service Level Objectives (SLOs)

Define and measure service quality with SLIs, SLOs, SLAs.

3. Eliminate Toil

Automate manual, repetitive work. Target <50% time on toil.

4. Monitoring & Alerting

Alert on symptoms (user-facing), not causes. Use golden signals.

5. Incident Response

Blameless postmortems, clear escalation, reduce MTTR.

6. Capacity Planning

Plan for growth, forecast demand, optimize resource usage.

SRE Workflow

  1. Define SLOs - What reliability do users need?
  2. Measure SLIs - Track service quality metrics
  3. Monitor error budget - How much budget consumed?
  4. Respond to incidents - Restore service quickly
  5. Conduct postmortems - Learn from failures
  6. Automate toil - Reduce manual work
  7. Plan capacity - Scale for growth

Google's Production Scale

SRE practices power:

  • Google Search: 8.5 billion searches/day
  • Gmail: 1.8 billion users
  • YouTube: 2 billion users, 1 billion hours/day
  • Google Maps: 1 billion users
  • 99.99%+ uptime across all services

Golden Signals (Google's 4 Key Metrics)

  1. Latency - Time to serve requests
  2. Traffic - Demand on system
  3. Errors - Failed requests
  4. Saturation - Resource utilization

Best Practices

  1. SLOs over SLAs - Internal targets stricter than external
  2. Error budget policy - Define consequences when budget exhausted
  3. Blameless culture - Learn from failures, don't blame
  4. Toil automation - Invest in eliminating repetitive work
  5. On-call sustainability - Max 25% on-call time, 50% ticket time

Related Skills


Last Updated: 2026-02-03 Expert: Alex Kim (Google SRE, 11 years) - Runs billion-user services

When not to use it

  • When the user aims for 100% uptime as a target.
  • When the user needs to focus on traditional operations or generic DevOps practices.
  • When the user wants to blame individuals for incidents.

Limitations

  • The skill advocates for embracing risk and not targeting 100% uptime.
  • It recommends spending less than 50% of time on toil.
  • It suggests a maximum of 25% on-call time and 50% ticket time for on-call sustainability.

How it compares

This skill provides a structured, software engineering-based approach to operations, emphasizing reliability metrics and automation, which differs from traditional operations or generic DevOps by prioritizing error budgets and blameless pos

Compared to similar skills

sre-practices side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
sre-practices (this skill)04moNo flagsAdvanced
observability-engineer124moNo flagsAdvanced
devops-troubleshooter14moNo flagsAdvanced
homelab-investigator04moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

Search skills

Search the agent skills registry