Learning Path Exam

Please answer the questions below to complete the test.

Question 1

When turning intent into a topology, what should a good infrastructure brief to an LLM include? (Select all that apply.)

Select all that apply

Functional requirements (what it must do)

Constraints (namespaces, images, limits)

Non-functional requirements (scale, security)

The model's training weights

Question 2

What role does Kanvas play when co-designing with an AI?

Select one answer

It trains the LLM on your cluster

It is the shared canvas and human-in-the-loop checkpoint for proposed designs

It replaces the need to review designs

It is a load generator

Question 3

A Meshery design (Kubernetes Manifest source) is fundamentally what?

Select one answer

A proprietary binary format

Valid Kubernetes YAML that Meshery can import and deploy

A Grafana dashboard

A container image manifest

Question 4

Which prompt instruction most improves the chance of importable design YAML?

Select one answer

'Be as verbose as possible'

'Output only valid YAML using these resource kinds; no commentary'

'Invent any apiVersion you like'

'Skip resource limits'

Question 5

Which are common mistakes in LLM-generated Kubernetes designs? (Select all that apply.)

Select all that apply

Deprecated or wrong apiVersion

Service selector not matching pod labels

Missing resource requests/limits

Placing resources in a Namespace

Question 6

In the agent deploy loop, what should happen before any change is applied to a cluster?

Select one answer

Apply first, review later

Preview with a dry-run/diff and get human approval

Delete the namespace

Disable validation

Question 7

What is the correct way to roll back a bad design change that is tracked in Git?

Select one answer

Edit production by hand and hope it sticks

git revert the change and re-deploy the previous known-good design

Delete the Git history

Turn off MeshSync

Question 8

What kind of error does Meshery's relationship validation catch that plain schema validation does not?

Select one answer

Spelling errors in comments

Semantic errors in how components relate to each other

Slow container image pulls

Git merge conflicts

Question 9

Which Kubernetes objects make good standing guardrails for AI-generated workloads?

Select all that apply

ResourceQuota

LimitRange

NetworkPolicy

ConfigMap

Question 10

Before merging AI-generated infrastructure, your review checklist should confirm correctness, security, resources, policy, and what else?

Type your answer below