Provider Test Patterns logo

Provider Test Patterns

Organization
hashicorp
provider-test-patterns

Terraform provider acceptance test patterns using terraform-plugin-testing with the Plugin Framework. Covers test structure, TestCase/TestStep fields, ConfigStateChecks with custom statecheck.StateCheck implementations, plan checks, CompareValue for cross-step assertions, config helpers, import testing with ImportStateKind, sweepers, and scenario patterns (basic, update, disappears, validation, regression), and ephemeral resource testing with the echoprovider package. Use when writing, reviewing, or debugging provider acceptance tests, including questions about statecheck, plancheck, TestCheckFunc, CheckDestroy, ExpectError, import state verification, ephemeral resources, or how to structure test files.

Overview

Publisherhashicorp
Repositoryagent-skills
Skill nameprovider-test-patterns
Stars
880
Forks
130
Bundled files
3
LicenseMPL-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by hashicorp on GitHub. Read the source before you install it.

Installation

Install the Provider Test Patterns AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/hashicorp/agent-skills/tree/main/plugins/terraform/skills/provider-test-patterns
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/hashicorp/agent-skills.git /tmp/agent-skills
mkdir -p .claude/skills
cp -r /tmp/agent-skills/plugins/terraform/skills/provider-test-patterns .claude/skills/provider-test-patterns
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Provider Test Patterns in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Provider Test Patterns on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Provider Test Patterns is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Provider Acceptance Test Patterns

Patterns for writing acceptance tests using terraform-plugin-testing with the Plugin Framework.

Source: HashiCorp Testing Patterns

References (load when needed):

  • references/checks.md — statecheck, plancheck, knownvalue types, tfjsonpath, comparers
  • references/sweepers.md — sweeper setup, TestMain, dependencies
  • references/ephemeral.md — ephemeral resource testing, echoprovider, multi-step patterns

Test Lifecycle

The framework runs each TestStep through: plan → apply → refresh → final plan. If the final plan shows a diff, the test fails (unless ExpectNonEmptyPlan is set). After all steps, destroy runs followed by CheckDestroy. This means every test automatically verifies that configurations apply cleanly and produce no drift — no assertions needed for that.


Test Function Structure

go
func TestAccExample_basic(t *testing.T) {
    var widget example.Widget
    rName := acctest.RandStringFromCharSet(10, acctest.CharSetAlphaNum)
    resourceName := "example_widget.test"

    resource.ParallelTest(t, resource.TestCase{
        PreCheck:                 func() { testAccPreCheck(t) },
        ProtoV6ProviderFactories: testAccProtoV6ProviderFactories,
        CheckDestroy:             testAccCheckExampleDestroy,
        Steps: []resource.TestStep{
            {
                Config: testAccExampleConfig_basic(rName),
                ConfigStateChecks: []statecheck.StateCheck{
                    stateCheckExampleExists(resourceName, &widget),
                    statecheck.ExpectKnownValue(resourceName,
                        tfjsonpath.New("name"), knownvalue.StringExact(rName)),
                    statecheck.ExpectKnownValue(resourceName,
                        tfjsonpath.New("id"), knownvalue.NotNull()),
                },
            },
        },
    })
}

Use resource.ParallelTest by default. Use resource.Test only when tests share state or cannot run concurrently.


Provider Factory

go
// provider_test.go — Plugin Framework with Protocol 6 (use Protocol5 variant if needed)
var testAccProtoV6ProviderFactories = map[string]func() (tfprotov6.ProviderServer, error){
    "example": providerserver.NewProtocol6WithError(New("test")()),
}

TestCase Fields

FieldPurpose
PreCheckfunc() — verify prerequisites (env vars, API access)
ProtoV6ProviderFactoriesPlugin Framework provider factories
CheckDestroyTestCheckFunc — verify resources destroyed after all steps
Steps[]TestStep — sequential test operations
TerraformVersionChecks[]tfversion.TerraformVersionCheck — gate by CLI version

TestStep Fields

Config Mode

FieldPurpose
ConfigInline HCL string to apply
ConfigStateChecks[]statecheck.StateCheck — modern assertions (preferred)
ConfigPlanChecksresource.ConfigPlanChecks{PreApply: []plancheck.PlanCheck{...}}
ExpectError*regexp.Regexp — expect failure matching pattern
ExpectNonEmptyPlanbool — expect non-empty plan after apply
PlanOnlybool — plan without applying
Destroybool — run destroy step
PreConfigfunc() — setup before step

Import Mode

FieldPurpose
ImportStatetrue to enable import mode
ImportStateVerifyVerify imported state matches prior state
ImportStateVerifyIgnore[]string — attributes to skip during verify
ImportStateKindresource.ImportBlockWithID — import block generation
ResourceNameResource address to import
ImportStateIdOverride the ID used for import

Check Functions

Modern: ConfigStateChecks (preferred)

Type-safe with aggregated error reporting. Compose built-in checks with custom statecheck.StateCheck implementations. See references/checks.md for full knownvalue types, tfjsonpath navigation, and comparers.

go
ConfigStateChecks: []statecheck.StateCheck{
    stateCheckExampleExists(resourceName, &widget),
    statecheck.ExpectKnownValue(resourceName,
        tfjsonpath.New("name"), knownvalue.StringExact("my-widget")),
    statecheck.ExpectKnownValue(resourceName,
        tfjsonpath.New("enabled"), knownvalue.Bool(true)),
    statecheck.ExpectKnownValue(resourceName,
        tfjsonpath.New("id"), knownvalue.NotNull()),
    statecheck.ExpectSensitiveValue(resourceName,
        tfjsonpath.New("api_key")),
},

Do not mix Check (legacy) and ConfigStateChecks in the same step.

Legacy: Check (for CheckDestroy and migration)

CheckDestroy on TestCase requires TestCheckFunc. The Check field on TestStep also accepts TestCheckFunc but prefer ConfigStateChecks for new tests.

go
Check: resource.ComposeAggregateTestCheckFunc(
    resource.TestCheckResourceAttr(name, "key", "expected"),
    resource.TestCheckResourceAttrSet(name, "id"),
    resource.TestCheckNoResourceAttr(name, "removed"),
    resource.TestMatchResourceAttr(name, "url", regexp.MustCompile(`^https://`)),
    resource.TestCheckResourceAttrPair(res1, "ref_id", res2, "id"),
),

ComposeAggregateTestCheckFunc reports all errors; ComposeTestCheckFunc fails fast on the first.


Config Helpers

Use numbered format verbs — %[1]q for quoted strings, %[1]s for raw:

go
func testAccExampleConfig_basic(rName string) string {
    return fmt.Sprintf(`
resource "example_widget" "test" {
  name = %[1]q
}
`, rName)
}

func testAccExampleConfig_full(rName, description string) string {
    return fmt.Sprintf(`
resource "example_widget" "test" {
  name        = %[1]q
  description = %[2]q
  enabled     = true
}
`, rName, description)
}

Scenario Patterns

Basic + Update (combine in one test — updates are supersets of basic)

go
Steps: []resource.TestStep{
    {
        Config: testAccExampleConfig_basic(rName),
        ConfigStateChecks: []statecheck.StateCheck{
            stateCheckExampleExists(resourceName, &widget),
            statecheck.ExpectKnownValue(resourceName,
                tfjsonpath.New("name"), knownvalue.StringExact(rName)),
        },
    },
    {
        Config: testAccExampleConfig_full(rName, "updated"),
        ConfigStateChecks: []statecheck.StateCheck{
            stateCheckExampleExists(resourceName, &widget),
            statecheck.ExpectKnownValue(resourceName,
                tfjsonpath.New("description"), knownvalue.StringExact("updated")),
        },
    },
},

Import

After a config step, verify import produces identical state. Use ImportStateKind for import block generation:

go
{
    ResourceName:      resourceName,
    ImportState:       true,
    ImportStateVerify: true,
    ImportStateKind:   resource.ImportBlockWithID,
},

Disappears (resource deleted externally)

go
{
    Config: testAccExampleConfig_basic(rName),
    ConfigStateChecks: []statecheck.StateCheck{
        stateCheckExampleExists(resourceName, &widget),
        stateCheckExampleDisappears(resourceName),
    },
    ExpectNonEmptyPlan: true,
},

Validation (expect error)

go
{
    Config:      testAccExampleConfig_invalidName(""),
    ExpectError: regexp.MustCompile(`name must not be empty`),
},

Regression (two-commit workflow)

A proper bug fix uses at least two commits: first commit the regression test (which fails, confirming the bug), then commit the fix (test passes). This lets reviewers independently verify the test reproduces the issue by checking out the first commit, then advancing to the fix.

Name and document regression tests to identify the issue they fix. Include a link to the original bug report when possible.

go
// TestAccExample_regressionGH1234 verifies fix for https://github.com/org/repo/issues/1234
func TestAccExample_regressionGH1234(t *testing.T) {
    rName := acctest.RandStringFromCharSet(10, acctest.CharSetAlphaNum)
    resourceName := "example_widget.test"

    resource.ParallelTest(t, resource.TestCase{
        PreCheck:                 func() { testAccPreCheck(t) },
        ProtoV6ProviderFactories: testAccProtoV6ProviderFactories,
        CheckDestroy:             testAccCheckExampleDestroy,
        Steps: []resource.TestStep{
            {
                // Reproduce the issue: this config triggered the bug
                Config: testAccExampleConfig_regressionGH1234(rName),
                ConfigStateChecks: []statecheck.StateCheck{
                    stateCheckExampleExists(resourceName, nil),
                    statecheck.ExpectKnownValue(resourceName,
                        tfjsonpath.New("computed_field"), knownvalue.NotNull()),
                },
            },
        },
    })
}

Helper Functions

Custom StateCheck: Exists

Implement statecheck.StateCheck for API existence verification. Separate the exists check into its own function for reuse across steps — the source recommends this as a design principle:

go
type exampleExistsCheck struct {
    resourceAddress string
    widget          *example.Widget
}

func (e exampleExistsCheck) CheckState(ctx context.Context, req statecheck.CheckStateRequest, resp *statecheck.CheckStateResponse) {
    r, err := stateResourceAtAddress(req.State, e.resourceAddress)
    if err != nil {
        resp.Error = err
        return
    }

    id, ok := r.AttributeValues["id"].(string)
    if !ok {
        resp.Error = fmt.Errorf("no id found for %s", e.resourceAddress)
        return
    }

    conn := testAccAPIClient()
    widget, err := conn.GetWidget(id)
    if err != nil {
        resp.Error = fmt.Errorf("%s not found via API: %w", e.resourceAddress, err)
        return
    }

    if e.widget != nil {
        *e.widget = *widget
    }
}

func stateCheckExampleExists(name string, widget *example.Widget) statecheck.StateCheck {
    return exampleExistsCheck{resourceAddress: name, widget: widget}
}

Custom StateCheck: Disappears

Delete a resource via API to simulate external deletion:

go
type exampleDisappearsCheck struct {
    resourceAddress string
}

func (e exampleDisappearsCheck) CheckState(ctx context.Context, req statecheck.CheckStateRequest, resp *statecheck.CheckStateResponse) {
    r, err := stateResourceAtAddress(req.State, e.resourceAddress)
    if err != nil {
        resp.Error = err
        return
    }

    id := r.AttributeValues["id"].(string)
    conn := testAccAPIClient()
    resp.Error = conn.DeleteWidget(id)
}

func stateCheckExampleDisappears(name string) statecheck.StateCheck {
    return exampleDisappearsCheck{resourceAddress: name}
}

State Resource Lookup (shared utility)

go
func stateResourceAtAddress(state *tfjson.State, address string) (*tfjson.StateResource, error) {
    if state == nil || state.Values == nil || state.Values.RootModule == nil {
        return nil, fmt.Errorf("no state available")
    }
    for _, r := range state.Values.RootModule.Resources {
        if r.Address == address {
            return r, nil
        }
    }
    return nil, fmt.Errorf("not found in state: %s", address)
}

Destroy Check (TestCheckFunc — required by CheckDestroy)

go
func testAccCheckExampleDestroy(s *terraform.State) error {
    conn := testAccAPIClient()
    for _, rs := range s.RootModule().Resources {
        if rs.Type != "example_widget" {
            continue
        }
        _, err := conn.GetWidget(rs.Primary.ID)
        if err == nil {
            return fmt.Errorf("widget %s still exists", rs.Primary.ID)
        }
        if !isNotFoundError(err) {
            return err
        }
    }
    return nil
}

PreCheck

go
func testAccPreCheck(t *testing.T) {
    t.Helper()
    if os.Getenv("EXAMPLE_API_KEY") == "" {
        t.Fatal("EXAMPLE_API_KEY must be set for acceptance tests")
    }
}

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Provider Test Patterns AI skill do?

Terraform provider acceptance test patterns using terraform-plugin-testing with the Plugin Framework. Covers test structure, TestCase/TestStep fields, ConfigStateChecks with custom statecheck.StateCheck implementations, plan checks, CompareValue for cross-step assertions, config helpers, import testing with ImportStateKind, sweepers, and scenario patterns (basic, update, disappears, validation, regression), and ephemeral resource testing with the echoprovider package. Use when writing, reviewing, or debugging provider acceptance tests, including questions about statecheck, plancheck, TestCh...

Why use Provider Test Patterns on TypingMind?

Because you install it once and use it with any model. Provider Test Patterns is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Provider Test Patterns in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/hashicorp/agent-skills/tree/main/plugins/terraform/skills/provider-test-patterns. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Provider Test Patterns?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Provider Test Patterns?

As many as you like. As long as a model supports skills, you can use Provider Test Patterns with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Provider Test Patterns AI skill free?

Yes. It is published on GitHub by hashicorp under the MPL-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇