Course Test

Question 1

A platform engineer needs to demonstrate to an auditor that a production infrastructure change was authorized before it was applied. Which artifact provides the clearest evidence of authorization?

Select one answer

The Meshery design YAML file as it exists in the repository at the time of audit

A Git merge commit from a branch that required pull request approval before merging

A MeshSync observation snapshot taken after the deployment

The output of mesheryctl system check run on the day of the change

Question 2

Which command imports a design file into Meshery for policy evaluation and versioning?

Select one answer

mesheryctl design apply -f designs/policy-guardrails.yaml

mesheryctl system start --design policy-guardrails.yaml

mesheryctl design import -f designs/policy-guardrails.yaml -s "Kubernetes Manifest"

mesheryctl perf apply -f designs/policy-guardrails.yaml

Question 3

In the five-layer operating model presented in this course, what is the role of the MCP server in the agent integration layer?

Select one answer

It stores the agent's reasoning history as a compliance artifact

It lets the agent bypass Meshery's policy checks to operate faster

It exposes Meshery capabilities to agents through a standard interface while enforcing the same policy and logging controls that apply to human operators

It replaces MeshSync for drift detection in agent-managed clusters

Question 4

What is the correct remediation path when an unmanaged resource is detected in a production cluster that represents an unauthorized out-of-band change?

Select one answer

Delete the resource immediately using kubectl without logging the action

Re-apply the Meshery design that represents the desired state, which produces a reconciliation attestation event

Disable MeshSync temporarily to stop generating drift alerts

Import the unmanaged resource into a new design without running policy validation

Question 5

Which of the following represent failure modes that degrade the AI-native operating model described in this course? Select all that apply.

Select all that apply

Making cluster changes directly with kubectl without an associated Meshery design

Removing human-in-the-loop approval gates from agent workflows to increase deployment speed

MeshSync becoming disconnected so drift events are not reported

Running mesheryctl system check before applying a new design

Question 6

What does 'continuous compliance' mean in the context of this course, and how does it differ from a point-in-time audit?

Type your answer below