TestChimp

Agent Assisted Test Execution

Agent Assisted
Test Execution

Scripted tests, with Agent step-in on demand. Code comments to drive QA Intelligence

Start Testing

Request Demo

Plain Playwright Scripts

TestChimp's SmartTests are standard Playwright scripts. This means:

Run in CI using the Playwright runner you already trust

Run in CI using standard Playwright runner

Fast & Scalable, unlike pure-agentic tests

Maintainable, with no vendor lock-in

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

ai.verify("No error messages");

ai.act("Close cookie modal");

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

AI-Native Steps When Needed

Use AI only where scripted steps become brittle with ai.act, ai.verify.

Easier authoring for complex flows

Less brittle tests

Retain the speed of scripts, with agentic flexibility

Speed of scripts, with agentic flexibility

QA Intelligence, Driven by Code Comments

QA Intelligence -
Driven by Code Comments

Link to scenarios right from code to enable coverage insights

Mark Screens and States for contextual bug reporting by Exploratory Agents

// @Screen: Dashboard

// @Scenario: checkout success

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

import { test, expect } from '@playwright/test';
import {ai} from 'ai-wright';
test('userMessagingFlow', async (
	{ page, browser, context }) => {
  
  var url = process.env.BASE_URL;
  var username = process.env.USERNAME;
  await page.goto(url);

  // @Screen: Login
  await page.getByLabel('Email').fill(username);

  await page.getByLabel('Password').fill('...');

  // Click on login button
  await page.getByRole('button').click();

  // @Screen: Dashboard
  await page.getByText('Messages').waitFor();

Author with ease

From a plain english scenario

Go from scenario description to executable test in one click

Learn more

From manual test steps

Record your manual tests with our Chrome Extension to generate automation scripts

Add To Chrome

Upload existing scripts

Sync your existing playwright test folder or upload your tests directly.

Learn more

Frequently Asked Questions

How are SmartTests different from plain Playwright scripts?

Traditional scripts can be brittle and flaky for complex interactions. TestChimp augments them with AI-native steps when needed, enabling agent step-in on demand, while keeping tests deterministic and maintainable by being script-first.

Why not rely entirely on agentic testing?

Pure agentic tests are slow (20ms for a script step vs 20s for an agentic step) and non-deterministic, making them not scalable. TestChimp provides a middle ground alternative of including AI-native steps in your existing scripts - retaining maintainability and speed.

How do AI-native steps work inside a test?

Simply say “ai.act(‘do X’)”; The agent steps in, looks at the screen, and decides the action to do based on the english description.

How are executions linked back to test plans?

Each test can reference scenarios in your test plans with a simple comment, allowing execution results, failures, and regressions to map directly back to planned coverage. This removes the need for manually maintained sheets, and makes traceability baked in to the code itself.

How does SmartTests help in Exploratory testing?

Exploratory agents are run using SmartTests as a GPS to guide the agent. This enables controlled navigation, measurable coverage and accurate attribution. // @screen annotations in the code helps agents to tag bugs found aligned with your app structure.

Is this suitable for large, CI-driven test suites?

Yes - SmartTests are just plain Playwright tests, so you can use your existing test runner. AI native step capability is available as an npm library that can be installed so your scripts can continue to run in CI.