Chaos Engineering — Development

This skill provides a disciplined framework for planning, running, and learning from chaos engineering experiments. It includes three Python tools for designing experiments, calculating blast radius, and generating postmortems, along with references on principles, attack taxonomy, and tooling. It enforces safety measures like abort criteria and steady-state metrics to prevent outages.

engineering claude-code

📂 View skill on GitHub ↗ 🚀 How to use

✨ What it does

Generates structured experiment plans with hypothesis, steady-state metric, blast radius, abort criteria, and rollback
Calculates expected affected users, error budget burn, and risk score (GREEN/YELLOW/RED) with recommendation
Produces structured postmortems that catch common failure modes like missing learnings and blame language
Provides a taxonomy of 7 attack types (latency, error, resource, partition, dependency, time, infrastructure) with tooli
Includes a tooling chooser with decision rules for Chaos Toolkit, Chaos Mesh, Litmus, Gremlin, and AWS FIS
Enforces the 4 Principles of Chaos Engineering plus a fifth: define abort criteria up front

🎯 When to use it

Planning a chaos experiment (what to break, where, when, how to abort)
Calculating blast radius before running an experiment
Reviewing an existing experiment plan for safety
Choosing a chaos tool (Chaos Toolkit, Chaos Mesh, Litmus, Gremlin, AWS FIS)
Writing a chaos experiment postmortem or running a Game Day exercise

🚀 How to use

Trigger: /chaos-experiment

Trigger the skill by mentioning chaos experiment, fault injection, gameday, resilience test, or using the /chaos-experiment slash command. Provide inputs like target service, hypothesis, attack type, duration, and blast radius. Example prompts:

Design a chaos experiment for checkout-svc with latency attack, hypothesis p99 < 500ms, duration 15 min, blast radius 5% of US traffic.
Calculate blast radius for 5% traffic share, 1M users, 15 min, baseline availability 0.999, expected impact 0.95.
Generate a postmortem from experiment plan and results log.

📄 Output: Markdown experiment plans, blast radius reports, and postmortems, plus reference documents and templates.

📦 Add this skill to Claude Code

# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills

# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/engineering/skills/chaos-engineering .claude/skills/chaos-engineering

Skill source: engineering/skills/chaos-engineering/SKILL.md

⚠️ Good to know

Not for incident response, threat hunting, performance load testing, or production debugging; requires clear hypothesis and abort criteria to be safe.

❓ FAQ

What are the 4 principles of chaos engineering?

The 4 principles are: build a hypothesis around steady-state behavior, vary real-world events, run experiments in production, and automate experiments to run continuously.

What is the risk score threshold for proceeding?

The blast radius calculator outputs a risk score: GREEN (<1% error budget) means PROCEED, YELLOW (1-10%) means REDUCE, and RED (>10%) means ABORT.

Which tool is best for a Kubernetes-only stack with OSS?

For a Kubernetes-only stack with OSS, Chaos Mesh or Litmus are recommended; Litmus has a larger experiment library.

🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.