Golang Stretchr Testify

Comprehensive guide to stretchr/testify for Golang testing. Covers assert, require, mock, and suite packages in depth. Use whenever writing tests with testif...

Security Scan

The skill is internally consistent: a testify-focused Go testing guide that reasonably requires Go and the gotests generator and contains only instruction content.

Purpose & Capability

Name/description match the requested binaries (go, gotests) and the install (github.com/cweill/gotests). Requiring gotests is proportional for a test-writing helper that may generate test stubs.

Instruction Scope

SKILL.md contains guidance limited to writing and reviewing tests with stretchr/testify; it does not instruct the agent to read unrelated files, environment variables, or exfiltrate data. The allowed tools (git, linters, gotests, WebFetch, etc.) are reasonable for a coding assistant.

Install Mechanism

Install uses 'go' to fetch github.com/cweill/gotests/...@latest which is a public Go package and common for developer tooling. This is expected for a test-generation helper but is a network fetch that will write a binary to disk (gotests); if you require stricter controls, review the upstream package source before installing.

Credentials

No environment variables, credentials, or config paths are requested. The skill does not ask for unrelated tokens or secrets.

Persistence & Privilege

always is false and the skill does not request elevated or persistent platform-wide privileges. Autonomous invocation is allowed by default but not combined with other concerning permissions.

Assessment

This skill is an instruction-only testify guide and appears coherent with its stated purpose. If you plan to install the provided gotests binary, review the github.com/cweill/gotests source (or vendor a vetted version) because 'go install' will download and build code from the network. Also confirm you’re comfortable with the agent using allowed tools (git, gotests, webfetch) and with autonomous invocation; no credentials or system-wide config access are requested.

Persona: You are a Go engineer who treats tests as executable specifications. You write tests to constrain behavior and make failures self-explanatory — not to hit coverage targets.

Modes:

• Write mode — adding new tests or mocks to a codebase.
• Review mode — auditing existing test code for testify misuse.

stretchr/testify

testify complements Go's testing package with readable assertions, mocks, and suites. It does not replace testing — always use *testing.T as the entry point.

This skill is not exhaustive. Please refer to library documentation and code examples for more informations. Context7 can help as a discoverability platform.

assert vs require

Both offer identical assertions. The difference is failure behavior:

• assert: records failure, continues — see all failures at once
• require: calls t.FailNow() — use for preconditions where continuing would panic or mislead

Use assert.New(t) / require.New(t) for readability. Name them is and must:

func TestParseConfig(t *testing.T) {
    is := assert.New(t)
    must := require.New(t)

    cfg, err := ParseConfig("testdata/valid.yaml")
    must.NoError(err)    // stop if parsing fails — cfg would be nil
    must.NotNil(cfg)

    is.Equal("production", cfg.Environment)
    is.Equal(8080, cfg.Port)
    is.True(cfg.TLS.Enabled)
}

Rule: require for preconditions (setup, error checks), assert for verifications. Never mix randomly.

Core Assertions

is := assert.New(t)

// Equality
is.Equal(expected, actual)              // DeepEqual + exact type
is.NotEqual(unexpected, actual)
is.EqualValues(expected, actual)        // converts to common type first
is.EqualExportedValues(expected, actual)

// Nil / Bool / Emptiness
is.Nil(obj)                  is.NotNil(obj)
is.True(cond)                is.False(cond)
is.Empty(collection)         is.NotEmpty(collection)
is.Len(collection, n)

// Contains (strings, slices, map keys)
is.Contains("hello world", "world")
is.Contains([]int{1, 2, 3}, 2)
is.Contains(map[string]int{"a": 1}, "a")

// Comparison
is.Greater(actual, threshold)     is.Less(actual, ceiling)
is.Positive(val)                  is.Negative(val)
is.Zero(val)

// Errors
is.Error(err)                     is.NoError(err)
is.ErrorIs(err, ErrNotFound)      // walks error chain
is.ErrorAs(err, &target)
is.ErrorContains(err, "not found")

// Type
is.IsType(&User{}, obj)
is.Implements((*io.Reader)(nil), obj)

Argument order: always (expected, actual) — swapping produces confusing diff output.

Advanced Assertions

is.ElementsMatch([]string{"b", "a", "c"}, result)             // unordered comparison
is.InDelta(3.14, computedPi, 0.01)                            // float tolerance
is.JSONEq(`{"name":"alice"}`, `{"name": "alice"}`)             // ignores whitespace/key order
is.WithinDuration(expected, actual, 5*time.Second)
is.Regexp(`^user-[a-f0-9]+$`, userID)

// Async polling
is.Eventually(func() bool {
    status, _ := client.GetJobStatus(jobID)
    return status == "completed"
}, 5*time.Second, 100*time.Millisecond)

// Async polling with rich assertions
is.EventuallyWithT(func(c *assert.CollectT) {
    resp, err := client.GetOrder(orderID)
    assert.NoError(c, err)
    assert.Equal(c, "shipped", resp.Status)
}, 10*time.Second, 500*time.Millisecond)

testify/mock

Mock interfaces to isolate the unit under test. Embed mock.Mock, implement methods with m.Called(), always verify with AssertExpectations(t).

Key matchers: mock.Anything, mock.AnythingOfType("T"), mock.MatchedBy(func). Call modifiers: .Once(), .Times(n), .Maybe(), .Run(func).

For defining mocks, argument matchers, call modifiers, return sequences, and verification, see Mock reference.

testify/suite

Suites group related tests with shared setup/teardown.

Lifecycle

SetupSuite()    → once before all tests
  SetupTest()   → before each test
    TestXxx()
  TearDownTest() → after each test
TearDownSuite() → once after all tests

Example

type TokenServiceSuite struct {
    suite.Suite
    store   *MockTokenStore
    service *TokenService
}

func (s *TokenServiceSuite) SetupTest() {
    s.store = new(MockTokenStore)
    s.service = NewTokenService(s.store)
}

func (s *TokenServiceSuite) TestGenerate_ReturnsValidToken() {
    s.store.On("Save", mock.Anything, mock.Anything).Return(nil)
    token, err := s.service.Generate("user-42")
    s.NoError(err)
    s.NotEmpty(token)
    s.store.AssertExpectations(s.T())
}

// Required launcher
func TestTokenServiceSuite(t *testing.T) {
    suite.Run(t, new(TokenServiceSuite))
}

Suite methods like s.Equal() behave like assert. For require: s.Require().NotNil(obj).

Common Mistakes

• Forgetting AssertExpectations(t) — mock expectations silently pass without verification
• is.Equal(ErrNotFound, err) — fails on wrapped errors. Use is.ErrorIs to walk the chain
• Swapped argument order — testify assumes (expected, actual). Swapping produces backwards diffs
• assert for guards — test continues after failure and panics on nil dereference. Use require
• Missing suite.Run() — without the launcher function, zero tests execute silently
• Comparing pointers — is.Equal(ptr1, ptr2) compares addresses. Dereference or use EqualExportedValues

Linters

Use testifylint to catch wrong argument order, assert/require misuse, and more. See samber/cc-skills-golang@golang-linter skill.

Cross-References

• → See samber/cc-skills-golang@golang-testing skill for general test patterns, table-driven tests, and CI
• → See samber/cc-skills-golang@golang-linter skill for testifylint configuration