How do you use test coverage meaningfully in Go — beyond just a percentage?

// Run coverage
// go test -cover ./...
// go test -coverprofile=cover.out ./...
// go tool cover -html=cover.out   # visual HTML report
// go tool cover -func=cover.out   # per-function percentages

// Useful: find UNCOVERED branches visually
// Red lines in cover -html are untested paths

// Coverage pragmas: mark code as intentionally not testable
func panicOnInvariantViolation(cond bool, msg string) {
    if !cond {
        // This code path requires a programming error to hit
        // It's fine to leave untested
        panic(msg) //nolint:gocritic
    }
}

// Testing edge cases found in coverage report
func parsePort(s string) (int, error) {
    n, err := strconv.Atoi(s)
    if err != nil {
        return 0, fmt.Errorf("port %q is not a number: %w", s, err) // branch 1
    }
    if n < 1 || n > 65535 {
        return 0, fmt.Errorf("port %d out of range [1, 65535]", n)   // branch 2
    }
    return n, nil // branch 3
}

// Table-driven test to hit all branches
func TestParsePort(t *testing.T) {
    tests := []struct {
        input   string
        want    int
        wantErr bool
    }{
        {"8080", 8080, false},    // happy path (branch 3)
        {"abc",  0,    true},     // not a number (branch 1)
        {"0",    0,    true},     // too low (branch 2)
        {"65536", 0,   true},     // too high (branch 2)
        {"1",    1,    false},    // boundary minimum
        {"65535", 65535, false},  // boundary maximum
    }
    // ... test loop
}

// CI enforcement:
// go test -coverprofile=cover.out ./...
// go tool cover -func=cover.out | tail -1 | awk '{print $3}' | grep -v '^[0-6][0-9]'