What problem does it solve?
Developing a direct Config Connector resource requires verifying that its controller behaves correctly against GCP, but running every test against real GCP is slow and costly. This Skill guides you through Phase 3 of direct resource development: building or extending a MockGCP service and aligning its behavior with real GCP responses.
Core Features & Use Cases
- Mock Service Implementation: Create or extend mock services under
mockgcp/mock<service_name>/ with CRUD methods for your resource kind, including proto generation and service registration.
- Log Alignment Workflow: Run
hack/compare-mock to diff mock output against recorded real GCP logs, then fix gaps using populate<ResourceKind>Defaults functions and scoped normalize.go rules.
- Ratcheting & Presubmit Compliance: Remove your resource from the re-reconciliation exclusion list in
tests/e2e/ratcheting.go and run fixture presubmits before committing.
- Use Case: You are adding a new
VertexAIDataset direct resource and need the mock aiplatform service to produce HTTP logs identical to real GCP so e2e fixture tests pass in CI.
Quick Start
Implement MockGCP support and align the mock logs for the VertexAIDataset direct resource using the existing e2e fixtures.