Assertions
An assertion is a short statement of fact about your service, typically one or two sentences and up to 1,000 characters. It describes something a failure mode assessment cannot determine from resource configuration alone. Examples include your expected traffic volume, whether the environment is production, the disaster recovery strategy you intend to use, or a component that runs on your infrastructure but is not an AWS resource.
Assertions belong to a service. Every failure mode assessment you run for that service uses the current set of assertions. Each assertion has one of two sources:
-
AI-generated (
AI_GENERATED) – Generated by the assessment engine during the first failure mode assessment of a service. These assertions make explicit what the engine inferred about your service so that you can review and correct them. The engine does not generate assertions that duplicate or contradict assertions you have already added. -
User-created (
USER) – Added by you, or created when you edit an AI-generated assertion. When you edit an AI-generated assertion, its source changes toUSERand the next generation of Resilience Hub never modifies it again.
Each assertion also records when it was created and last updated. You can view, add, edit, and delete assertions in the console or with the AWS CLI at any time, before or after running an assessment.
