node-testing

name	node-testing
description	Write and run tests using Node.js 25's built-in test runner (node:test). Zero dependencies — includes describe/it, mocking, coverage, expectFailure, rerun-failures, and TypeScript type stripping. Use when: writing unit/integration tests, running test suites, checking coverage, mocking functions, or testing .ts files directly.
metadata	{"openclaw":{"emoji":"🧪","requires":{"bins":["node"]},"compat":"Node.js >= 25.x (24.x partially supported; expectFailure requires 25.5+)"}}

Node.js 25 Built-In Test Runner

Zero-dependency testing with node:test. No Jest, no Mocha, no Vitest — everything is built in.

When to Use

✅ USE this skill when:

• Writing unit tests or integration tests for Node.js projects
• Running test suites for CLI tools, APIs, libraries
• Checking code coverage on a project
• Mocking functions, methods, or timers in tests
• Testing TypeScript files directly (type stripping)
• User says: "run the tests", "write tests for X", "test coverage", "mock something"

❌ DON'T use this skill when:

• Browser/frontend testing (use Playwright, Vitest browser mode)
• React component testing (use Vitest + Testing Library)
• Projects locked to Jest/Vitest (respect existing config)

Quick Start

# Run all tests (auto-discovers *.test.js, *_test.js, etc.)
node --test

# Run specific file
node --test src/math.test.js

# Run with coverage
node --test --experimental-test-coverage

Core APIs

describe / it (BDD-style)

import { describe, it, beforeEach, afterEach } from 'node:test';
import assert from 'node:assert';

describe('Calculator', () => {
  let calc;

  beforeEach(() => {
    calc = { add: (a, b) => a + b };
  });

  it('should add two numbers', () => {
    assert.strictEqual(calc.add(2, 3), 5);
  });

  it('should handle negatives', () => {
    assert.strictEqual(calc.add(-1, 1), 0);
  });

  describe('nested suite', () => {
    it('works inside describe', () => {
      assert.ok(true);
    });
  });
});

Aliases: describe = suite, it = test. Use whichever you prefer.

test() (functional style)

import test from 'node:test';
import assert from 'node:assert';

test('synchronous', () => {
  assert.strictEqual(1, 1);
});

test('async', async () => {
  const result = await fetch('https://example.com');
  assert.ok(result.ok);
});

test('callback', (t, done) => {
  setTimeout(done, 100);
});

Subtests

test('parent', async (t) => {
  await t.test('child 1', () => { /* ... */ });
  await t.test('child 2', () => { /* ... */ });
});

Built-In Mocking

No jest.fn() needed. Mocking is built into node:test.

Mock Functions (Spy)

import { mock, test } from 'node:test';
import assert from 'node:assert';

test('spy on a function', () => {
  const fn = mock.fn((a, b) => a + b);

  assert.strictEqual(fn(3, 4), 7);
  assert.strictEqual(fn.mock.callCount(), 1);

  const call = fn.mock.calls[0];
  assert.deepStrictEqual(call.arguments, [3, 4]);
  assert.strictEqual(call.result, 7);

  mock.reset(); // Reset all globally tracked mocks
});

Mock Methods (auto-restore via test context)

test('mock object method', (t) => {
  const obj = {
    value: 5,
    add(a) { return this.value + a; }
  };

  t.mock.method(obj, 'add');
  obj.add(3);

  assert.strictEqual(obj.add.mock.callCount(), 1);
  assert.deepStrictEqual(obj.add.mock.calls[0].arguments, [3]);

  // Auto-restored when test finishes — no manual cleanup needed
});

Mock Timers

import { mock, test } from 'node:test';
import assert from 'node:assert';

test('mock setTimeout', () => {
  const fn = mock.fn();

  mock.timers.enable({ apis: ['setTimeout'] });
  setTimeout(fn, 9999);
  assert.strictEqual(fn.mock.callCount(), 0);

  mock.timers.tick(9999); // Fast-forward time
  assert.strictEqual(fn.mock.callCount(), 1);

  mock.timers.reset();
  mock.reset();
});

Supported timer APIs: setTimeout, setInterval, Date.now(), Date constructor, setImmediate.

Mock Modules (getters/setters)

test('mock getter', (t) => {
  const obj = { get val() { return 'real'; } };
  t.mock.getter(obj, 'val', () => 'mocked');

  assert.strictEqual(obj.val, 'mocked');
});

Test Options

Skipping Tests

// Skip via option
it('skipped', { skip: true }, () => { /* never runs */ });
it('skipped with reason', { skip: 'bug #123' }, () => {});

// Skip via method
it('also skipped', (t) => {
  t.skip();           // no message
  t.skip('reason');   // with message
});

// Skip entire suite
describe.skip('skipped suite', () => { /* ... */ });

TODO Tests (expected failures, v25.5+)

TODO tests execute but don't fail the suite. Use for pending work.

it('not yet implemented', { todo: true }, () => {
  throw new Error('this does NOT fail the test');
});

it('with message', { todo: 'needs refactoring' }, () => {});

expectFailure — Mark Known Bugs (v25.5+)

🔄 Reversed logic: The test must throw to pass. If it doesn't throw (i.e., it runs clean), it fails.

Perfect for tracking known bugs: when the bug is fixed, the test starts failing — alerting you to remove the flag.

// Basic: any failure = pass
it.expectFailure('known bug: fails intermittently', () => {
  assert.strictEqual(buggyFunction(), expected);
});

// With options object
it('known bug', { expectFailure: true }, () => {
  assert.strictEqual(buggyFunction(), expected);
});

// With reason string
it('known bug', { expectFailure: 'feature not implemented' }, () => {
  assert.strictEqual(buggyFunction(), expected);
});

// Match specific error (regex)
it('fails with specific error', { expectFailure: /timeout/i }, () => {
  throw new Error('Connection timeout');
});

// Match specific error (object shape)
it('error code match', { expectFailure: { code: 'ERR_EXPECTED' } }, () => {
  const err = new Error('boom');
  err.code = 'ERR_EXPECTED';
  throw err; // matches → test passes
});

// Both reason AND specific error
it('specific failure', {
  expectFailure: { label: 'known bug', match: /error msg/ }
}, () => {
  throw new Error('error msg here');
});

Priority: skip > todo > expectFailure. If a test has both skip and expectFailure, it's skipped.

Only Tests (focused runs)

# Must pass --test-only flag
node --test --test-only

it.only('only this runs', () => { /* ... */ });
describe.only('only this suite', () => { /* ... */ });

Filtering by Name Pattern

# Run tests matching regex
node --test --test-name-pattern="should add"
node --test --test-name-pattern="/should add/i"

# Skip tests matching regex
node --test --test-skip-pattern="slow"

Code Coverage

# Basic coverage (spec reporter shows summary)
node --test --experimental-test-coverage

# Generate lcov report
node --test --experimental-test-coverage \
  --test-reporter=lcov \
  --test-reporter-destination=coverage.info

# Include node_modules in coverage
node --test --experimental-test-coverage \
  --test-coverage-include="**/*.js"

# Exclude specific paths
node --test --experimental-test-coverage \
  --test-coverage-exclude="**/migrations/**"

Inline Coverage Control

/* node:coverage disable */
// Lines here are ignored for coverage
/* node:coverage enable */

/* node:coverage ignore next */
if (neverReached) { console.log('nope'); }

/* node:coverage ignore next 3 */
if (neverReached) {
  console.log('line 1');
  console.log('line 2');
}

--test-rerun-failures — Only Rerun Failed Tests

Instead of re-running the entire suite, persist state and only re-run what failed:

# First run: saves state to a file
node --test --test-rerun-failures ./test-state.json

# Second run: only re-runs tests that failed
node --test --test-rerun-failures ./test-state.json

# Third run: keeps narrowing down
node --test --test-rerun-failures ./test-state.json

The state file is JSON tracking which tests passed on which attempt. Each subsequent run skips passing tests and only re-executes failures. Great for CI flaky-test retry loops.

TypeScript — Type Stripping (Zero Config)

Node.js 25 has built-in type stripping. No ts-node, no tsx, no transpile step.

# Run .ts test files directly (type stripping is ON by default in Node 25)
node --test src/utils.test.ts

# Explicitly disable type stripping
node --test --no-strip-types src/utils.test.ts

Auto-discovery of .ts test files (when type stripping is enabled):

• *.test.{cts,mts,ts}
• *-test.{cts,mts,ts}
• *_test.{cts,mts,ts}
• test-*.{cts,mts,ts}
• test.{cts,mts,ts}
• test/**/*.{cts,mts,ts}

Limitations: Type stripping removes types but does NOT check them. Run tsc --noEmit separately for type checking.

// math.test.ts — runs directly with node --test
import { describe, it } from 'node:test';
import assert from 'node:assert';

function add(a: number, b: number): number {
  return a + b;
}

describe('add', () => {
  it('sums correctly', () => {
    assert.strictEqual(add(2, 3), 5);
  });
});

Watch Mode

node --test --watch

Automatically re-runs tests when files change. Good for TDD workflow.

Global Setup / Teardown (v24+)

// setup.js
export async function globalSetup() {
  console.log('Starting DB...');
  // await startDatabase();
}

export async function globalTeardown() {
  console.log('Stopping DB...');
  // await stopDatabase();
}

node --test --test-global-setup=./setup.js

Reporters

# Spec (default, human-readable)
node --test --test-reporter=spec

# TAP (machine-readable)
node --test --test-reporter=tap

# LCOV (for coverage tools)
node --test --experimental-test-coverage --test-reporter=lcov --test-reporter-destination=lcov.info

# Dot (minimal)
node --test --test-reporter=dot

Key Flags Reference

Flag	Description
--test	Enable test runner, auto-discover test files
--test-concurrency=N	Max parallel child processes (default: CPU count)
--test-name-pattern=REGEX	Run only tests matching pattern
--test-skip-pattern=REGEX	Skip tests matching pattern
--test-only	Only run tests with { only: true }
--test-rerun-failures=FILE	Persist state, only rerun failed tests
--test-reporter=TYPE	Output format: spec, tap, dot, lcov
--test-reporter-destination=FILE	Reporter output destination
--test-global-setup=FILE	Global setup/teardown module
--experimental-test-coverage	Enable code coverage collection
--test-coverage-include=GLOB	Coverage include pattern
--test-coverage-exclude=GLOB	Coverage exclude pattern
--watch	Re-run tests on file changes
--test-force-exit	Force exit after tests complete
--experimental-strip-types	Enable TS type stripping (default in Node 25)
--no-strip-types	Disable type stripping