testing-go

name	testing-go
description	Go testing patterns for MCP Toolkit. Use this when writing or reviewing tests — table-driven tests, mocking, coverage thresholds, and E2E testing.

Skill: Go Testing Patterns

Table-Driven Tests

Always prefer table-driven tests with testify:

func TestSomething(t *testing.T) {
    tests := []struct {
        name    string
        input   string
        want    string
        wantErr bool
    }{
        {"valid input", "input", "output", false},
        {"empty input returns error", "", "", true},
    }
    for _, tt := range tests {
        t.Run(tt.name, func(t *testing.T) {
            got, err := Function(tt.input)
            if tt.wantErr {
                assert.Error(t, err)
                return
            }
            assert.NoError(t, err)
            assert.Equal(t, tt.want, got)
        })
    }
}

Mocking

Mock at the interface boundary, not the implementation:

type mockRuntime struct{ available bool }

func (m *mockRuntime) Available() bool { return m.available }
func (m *mockRuntime) Name() string    { return "mock" }

Keep mocks in the same _test.go file that uses them.

Test Naming

Use TestUnit_Scenario_ExpectedOutcome:

func TestDetectRuntime_NoRuntimes_ReturnsError(t *testing.T) { ... }
func TestDetectRuntime_DockerAvailable_ReturnsDocker(t *testing.T) { ... }

Coverage

• Minimum 90% coverage (enforced by make coverage-check).
• Mock external dependencies (filesystem, container runtime, editor paths).
• Test error cases, not just happy paths.

E2E Tests

E2E tests live in test/e2e/ and exercise the compiled binary:

func TestCLI_Version_PrintsVersion(t *testing.T) {
    out, err := exec.Command(binary, "version").CombinedOutput()
    assert.NoError(t, err)
    assert.Contains(t, string(out), "mcp-toolkit")
}

Corner Cases Checklist

Before marking a feature tested, verify:

• Nil/zero/empty input?
• Boundary values (first, last, at-limit)?
• Dependency fails mid-operation?
• Missing required configuration?
• File permission errors?