菜单
Create test fixture dependencies (keys, issuers, targets, secrets) via SDK API calls in test setup, not via Terraform depends_on. Use when writing or reviewing acceptance tests that need pre-existing items, or when a test fails because a Terraform resource references an item that doesn't exist.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Guides adding a new Terraform resource or data source end-to-end — SDK update, schema, CRUD, provider registration, tests, and formatting. Use when creating a new resource_*.go or data_source_*.go file, or when the user asks to add a new Terraform resource.
Correct field to use for item descriptions in Read functions — ItemMetadata for DescribeItem, Description for specific Get endpoints. Use when implementing or reviewing d.Set("description", ...) in resource read functions.
Pattern for Terraform resource import functions — dedicated per-file import func that delegates to Read. Use when adding an Importer to a resource, implementing terraform import support, or reviewing import functions.
Guides updating an existing Terraform resource or data source with new fields or functionality — schema changes, CRUD updates, backward compatibility, and docs regeneration. Use when adding fields to an existing resource_*.go or data_source_*.go file.
| name | test-fixture-setup |
| description | Create test fixture dependencies (keys, issuers, targets, secrets) via SDK API calls in test setup, not via Terraform depends_on. Use when writing or reviewing acceptance tests that need pre-existing items, or when a test fails because a Terraform resource references an item that doesn't exist. |
When an acceptance test resource depends on a pre-existing item (e.g. a gateway config resource that references an SSH cert issuer), create that item via the SDK API in the test function body, not as a Terraform resource with depends_on.
Why: Terraform resources inside test configs can fail with ordering issues, and mixing infrastructure setup with the resource under test makes failures harder to diagnose. API-created fixtures are deterministic and independent of the Terraform run.
func TestMyResource(t *testing.T) {
testutils.SkipIfNoGateway(t)
// 1. Create fixtures via API
keyPath := testPath("my_key")
testutils.CreateDfcKey(t, keyPath)
t.Cleanup(func() {
testutils.DeleteItem(t, keyPath)
})
issuerPath := testPath("my_issuer")
testutils.CreateSshCertIssuer(t, keyPath, issuerPath, "test")
t.Cleanup(func() {
testutils.DeleteItem(t, issuerPath)
})
// 2. Reference fixtures by path in Terraform config
config := fmt.Sprintf(`
resource "akeyless_my_resource" "%v" {
some_item_ref = "%v"
other_field = "value1"
}
`, name, issuerPath)
configUpdate := fmt.Sprintf(`
resource "akeyless_my_resource" "%v" {
some_item_ref = "%v"
other_field = "value2"
}
`, name, issuerPath)
testutils.TestGatewayConfigResource(t, providerFactories, config, configUpdate)
}
All in akeyless/tests/testutils/testutils.go:
| Helper | Creates | Requires |
|---|---|---|
CreateDfcKey(t, name) | DFC key (RSA1024) | — |
CreateProtectionKey(t, name) | AES128-GCM key | — |
CreateSshCertIssuer(t, keyName, issuerName, users) | SSH cert issuer | DFC key |
CreatePkiCertIssuer(t, keyName, issuerName, destPath, cn, uriSan) | PKI cert issuer | DFC key |
CreateCertificate(t, certName, certB64, keyB64) | Certificate item | Generated cert/key pair |
CreateSecret(t, *TestSecret) | Static secret | — |
DeleteItem(t, path) | — | Always defer after create |
DeleteItemIfExists(t, path) | — | Silent if missing |
Always defer testutils.DeleteItem(t, path) immediately after each create call, before the next create. This ensures cleanup runs in reverse order even if a later create fails.