new-step

name	new-step
description	Scaffolds a new KEB provisioning/deprovisioning/update step with implementation and test file.

new-step

Scaffold a new KEB process step (provisioning, deprovisioning, or update) with its implementation file and test file.

Usage

/new-step <StepName> [provisioning|deprovisioning|update]

Examples:

• /new-step ValidateQuota provisioning
• /new-step ArchiveAuditLog deprovisioning
• /new-step SyncLabels update

If the operation type is omitted, ask the user which one they want.

---

What to do

Given <StepName> and the operation type, produce two files:

1. Implementation file

Path: internal/process/<operation_type>/<snake_case_name>_step.go

Use this template (replace {{StepName}}, {{snake_case_name}}, {{Pascal_Snake_name}}, {{package}}):

• {{StepName}} — PascalCase struct name, e.g. ValidateQuota
• {{snake_case_name}} — file name part, e.g. validate_quota
• {{Pascal_Snake_name}} — Name() return value: PascalCase words joined by underscores, e.g. Validate_Quota
• {{package}} — Go package name (provisioning, deprovisioning, or update)

package {{package}}

import (
	"log/slog"
	"time"

	"github.com/kyma-project/kyma-environment-broker/internal"
	kebError "github.com/kyma-project/kyma-environment-broker/internal/error"
	"github.com/kyma-project/kyma-environment-broker/internal/process"
	"github.com/kyma-project/kyma-environment-broker/internal/storage"
)

type {{StepName}}Step struct {
	operationManager *process.OperationManager
}

func New{{StepName}}Step(os storage.Operations) *{{StepName}}Step {
	step := &{{StepName}}Step{}
	step.operationManager = process.NewOperationManager(os, step.Name(), kebError.KEBDependency)
	return step
}

func (s *{{StepName}}Step) Name() string {
	return "{{Pascal_Snake_name}}"
}

func (s *{{StepName}}Step) Run(operation internal.Operation, log *slog.Logger) (internal.Operation, time.Duration, error) {
	// TODO: implement
	return operation, 0, nil
}

Key rules:

• kebError dependency: use KEBDependency as default; change to InfrastructureManagerDependency, LifeCycleManagerDependency, etc. if the step calls an external system — ask the user if unsure.
• Add only the dependencies actually needed (storage, k8sClient, etc.). Don't add fields speculatively.
• The package name matches the directory: provisioning, deprovisioning, or update.

2. Test file

Path: internal/process/<operation_type>/<snake_case_name>_step_test.go

Use this template:

Use the fixture function that matches the operation type:

• provisioning → fixture.FixProvisioningOperation("op-id", "instance-id")
• deprovisioning → fixture.FixDeprovisioningOperation("op-id", "instance-id")
• update → fixture.FixUpdatingOperation("op-id", "instance-id")

package {{package}}

import (
	"log/slog"
	"testing"

	"github.com/kyma-project/kyma-environment-broker/internal/fixture"
	"github.com/kyma-project/kyma-environment-broker/internal/storage"
	"github.com/stretchr/testify/assert"
	"github.com/stretchr/testify/require"
)

func Test{{StepName}}Step_HappyPath(t *testing.T) {
	// given
	st := storage.NewMemoryStorage()
	step := New{{StepName}}Step(st.Operations())

	op := fixture.Fix{{OperationType}}Operation("op-id", "instance-id")
	require.NoError(t, st.Operations().InsertOperation(op))

	// when
	result, retry, err := step.Run(op, slog.Default())

	// then
	assert.NoError(t, err)
	assert.Zero(t, retry)
	_ = result
}

Key rules:

• Use fixture.FixProvisioningOperation / fixture.FixDeprovisioningOperation / fixture.FixInstance from internal/fixture/.
• Use storage.NewMemoryStorage() — never PostgreSQL in unit tests.
• Use slog.Default() for the logger. If the package already defines a fixLogger() helper (check other _test.go files in the same package), use that instead — but do not redefine it.
• Use require for setup (fatal on failure), assert for outcome checks.

3. Registration reminder

After creating the files, remind the user:

Next step: Register the new step in cmd/broker/<operation_type>.go by adding an entry to the <operation_type>Steps slice:

{
    step: <package>.New{{StepName}}Step(db.Operations()),
},

Add it at the position in the pipeline where it logically belongs. If you want a conditional step, add a condition field (provisioning only).

---

Conventions to follow

• Step names use PascalCase for the struct; Name() returns the step name with words separated by underscores (e.g. struct ValidateQuotaStep → Name() returns "Validate_Quota", struct ApplyKymaStep → Name() returns "Apply_Kyma").
• Do not add error handling for impossible paths.
• Do not add comments unless the logic is non-obvious.
• Match the style of existing steps in the same package — read one before generating.