---
title: "Controlled failure-experiment designer: review, role & definition · Undominated.ai"
canonical: https://undominated.ai/agents/voltagent-chaos-engineer/
description: "Frames failure experiments around a steady-state hypothesis, bounded impact, observable recovery and explicit stop controls."
---

# Controlled failure-experiment designer: review, role & definition · Undominated.ai

> Frames failure experiments around a steady-state hypothesis, bounded impact, observable recovery and explicit stop controls.

[← Explore all agents](/agents/)

RESILIENCE TESTING / VoltAgent

# Controlled failure-experiment designer

Frames failure experiments around a steady-state hypothesis, bounded impact, observable recovery and explicit stop controls.

 [Use this definition ↓](#setup)[Original source ↗](https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/categories/04-quality-security/chaos-engineer.md)

SOURCE REVIEW

 Reviewed 2026-10-14
 Evidence 3 linked sources
 Publisher VoltAgent
 Licence [MIT ↗](https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/LICENSE)
 Revision 408d3c9eee2d
 [Read what was—and wasn’t—checked ↓](#review)

“Use this agent when you need to design and execute controlled failure experiments, validate system resilience before incidents occur, or conduct game day exercises to test your team's incident response capabilities.”

 [VoltAgent · upstream description ↗](https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/categories/04-quality-security/chaos-engineer.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - The workflow begins in non-production and expands scope gradually rather than beginning with combined failures.
- It includes human response and communication observations as well as infrastructure/service failures.
- The profile designs failure experiments that start outside production, with a limited blast radius and a rollback path.
- Its checklist requires a stated steady state and a hypothesis before an experiment, and it says customer impact is not acceptable.

### A good fit for

 - Provide system dependencies, SLOs, incident history, recovery procedures and a specifically authorized experiment environment.
- Receive an experiment plan with one variable, blast radius, rollback/kill switch, metrics and a learning/runbook update.
- Writing a game-day plan that starts in a non-production environment.
- Checking that an experiment proposal names its blast radius, its rollback, and its abort path.

### Weigh up before choosing

 - The profile provides no executable safety controller or permission boundary; the host must enforce the approved failure targets and abort mechanism.
- Rollback timing, zero impact and MTTR/resilience improvements in the examples are aspirations, not demonstrated guarantees.
- The source selects the sonnet model alias; confirm it exists in the configured host.
- Source-only listing: inspect the upstream context and referenced dependencies before adapting this definition.
- Tools include Write, Edit, and Bash, so the host could change files and run commands. This review did not inject a failure.
- The rollback time and the experiment counts in the file are checklist and template text, not a measured drill.
- The file lists manual kill switches. It does not implement one.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - categories/04-quality-security/chaos-engineer.md (complete definition)
- LICENSE (applicable redistribution terms)
- Host custom-agent documentation; repository installation guidance where applicable

### Our findings

 - The workflow begins in non-production and expands scope gradually rather than beginning with combined failures.
- It includes human response and communication observations as well as infrastructure/service failures.
- The checklist requires a steady state, a hypothesis, a controlled blast radius, and no customer impact.
- Experiment patterns say to begin in non-production, test one variable, and keep a rollback path. Blast-radius controls include a manual kill switch.
- Declared tools are Read, Write, Edit, Bash, Glob, and Grep.
- The sample progress JSON hard-codes experiments, failures, and an MTTR figure, and the checklist states a rollback time. None of those figures was measured here, and the file does not run an experiment by being read.

### Not established by this review

 - No agent was installed or invoked. Host/model compatibility, task accuracy and runtime behaviour were not tested.
- The resource was not executed and its tools were not called.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Use the role in your project.

[Upstream setup instructions ↗](https://code.claude.com/docs/en/sub-agents.md)
 - Read the pinned definition and MIT licence; review the host permissions before use.
- Place the reviewed definition in .claude/agents/, supply the system context directly and constrain file/shell/cloud permissions to the agreed diagnostic or test scope.
- Provide system dependencies, SLOs, incident history, recovery procedures and a specifically authorized experiment environment.

### Before you start

 - Claude Code with the selected model available
- The specific infrastructure/toolchain and a bounded non-production validation environment

### Compatibility

Claude Code

### Implementation

Markdown

### Host-mediated repository and tool access

 - Declared tools: Read, Write, Edit, Bash, Glob, Grep. The host enforces tool access; shell execution can modify files or contact services.

### Cost model

The definition is reusable under its stated licence. Host/model usage, execution environments and connected services can incur separate charges.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Complete upstream definition at reviewed commit ↗](https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/categories/04-quality-security/chaos-engineer.md) Checked 2026-10-14 https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/categories/04-quality-security/chaos-engineer.md Supports: Workflow, inputs and outputs, Declared tools and constraints, Selection and limitations
- [Applicable upstream redistribution licence ↗](https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/LICENSE) Checked 2026-10-14 https://github.com/VoltAgent/awesome-claude-code-subagents/blob/408d3c9eee2d0c75c12a3abb5de29e53c2ca6b9e/LICENSE Supports: Licence and attribution requirements
- [Host custom-agent configuration ↗](https://code.claude.com/docs/en/sub-agents.md) Checked 2026-10-14 https://code.claude.com/docs/en/sub-agents.md Supports: Host setup and permission defaults

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Database reliability administrator ↗ Plans database availability and maintenance around tested recovery, observed resource limits and documented operational procedures.](/agents/wshobson-database-admin/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Find reusable instructions](/skills/)
- [Inspect connections](/mcp-servers/)
- [Build a toolkit](/tools/)
- [Choose the model](/compare/)
