Use this template for deterministic resilience tests or chaos experiments when failure and recovery behavior matters.

Production chaos always requires explicit human authorization.

• Project:
• Phase, story, or feature:
• Date:
• Owner:
• Status: Draft | Ready | Complete | Blocked
• Experiment type: deterministic-resilience-test | chaos-experiment
• Linked behavior spec:
• Linked phase or feature plan:

• User-visible behavior:
• Operator-visible behavior:
• Data integrity condition:
• Telemetry:
• Baseline command or dashboard:

• If:
• When:
• Then:

• Target:
• Fault type:
• Scope:
• Duration:
• Traffic level:
• Environment:

Fault types may include process termination, dependency outage, added latency, packet loss, connection reset, network partition, CPU or memory pressure, connection-pool exhaustion, disk pressure, queue backlog, duplicate messages, delayed messages, out-of-order messages, and zone or region failure.

• Abort conditions:
• Data-loss prevention:
• Customer-impact limit:
• Cost limit:
• Production authorization:

• User-visible behavior:
• Operator-visible behavior:
• Retry or fallback:
• Recovery objective:
• Graceful degradation:

• Commands:
• Logs:
• Metrics:
• Traces:
• Screenshots:
• Data-integrity result:

• Observed result:
• Hypothesis result: confirmed | rejected | inconclusive
• Remediation:
• Residual risk:
• Follow-up: