Puppeteer Browser Automation Runner

What is Puppeteer Browser Automation Runner?

Puppeteer Browser Automation Runner executes safe, declarative browser workflows in the Apify cloud using Puppeteer and Chrome.

Instead of writing and hosting a complete Puppeteer application, define an ordered list of actions such as navigation, clicks, form input, extraction, assertions, waits, and screenshots. The Actor executes the workflow and stores one structured Dataset row per step.

It is designed for browser automation, repetitive web workflows, smoke monitoring, lightweight extraction, and integration into scheduled or API-driven pipelines.

Supported actions

• goto - navigate to the target URL or another public HTTP(S) URL
• waitForSelector - wait until a CSS selector exists
• click - click an element
• type - type text into a field
• select - select an option in a <select> element
• extractText - return an element's text content
• extractAttribute - return an HTML attribute such as href or src
• wait - pause the workflow for a specified number of milliseconds
• screenshot - capture a screenshot and save it to the run Key-Value Store
• textContains - assert that visible page text contains a value
• urlContains - assert that the current URL contains a value
• titleContains - assert that the page title contains a value

Example input

{
  "targetUrl": "https://example.com",
  "workflowName": "Example.com workflow",
  "actions": [
    {
      "type": "goto",
      "label": "Open target page"
    },
    {
      "type": "waitForSelector",
      "label": "Wait for heading",
      "selector": "h1"
    },
    {
      "type": "extractText",
      "label": "Extract heading",
      "selector": "h1"
    },
    {
      "type": "textContains",
      "label": "Check expected text",
      "value": "Example Domain"
    },
    {
      "type": "screenshot",
      "label": "Capture page"
    }
  ]
}

Example output

Each executed step is stored as a separate Dataset item:

{
  "workflowName": "Example.com workflow",
  "targetUrl": "https://example.com",
  "step": 3,
  "action": "extractText",
  "label": "Extract heading",
  "passed": true,
  "currentUrl": "https://example.com/",
  "value": "Example Domain",
  "error": null,
  "screenshotUrl": null
}

Failed steps are also stored, with passed: false and an error message. Depending on the input settings, the workflow can stop on the first failure or continue with later steps.

Screenshots and artifacts

Explicit screenshot actions always save a PNG to the run's default Key-Value Store and return its URL in the Dataset.

You can also configure automatic screenshots:

• on failed steps
• after every step
• only for explicit screenshot actions

Use cases

Browser workflow automation

Automate predictable sequences of navigation, clicks, form entry, and page interactions without deploying your own browser infrastructure.

Website monitoring

Schedule workflows to verify important text, URLs, selectors, or page titles and inspect failure screenshots.

Lightweight web extraction

Extract text or attributes from JavaScript-rendered websites and send structured results to another system through the Apify API.

Repetitive QA workflows

Automate browser checks that do not require a full Cypress or Playwright test suite.

Data and AI pipelines

Use the Actor as a browser-action component inside scheduled workflows, backend services, or AI-driven pipelines.

API and scheduling

The Actor can be run from the Apify Console, the Apify API, schedules, webhooks, and integrations.

Because every workflow step is written to the default Dataset, downstream systems can consume results as JSON, CSV, Excel, XML, or through the Dataset API.

Safety model

This Actor intentionally uses declarative actions instead of arbitrary user-supplied JavaScript.

For security, requests to localhost, loopback addresses, link-local addresses, and common private-network IP ranges are blocked. Only public HTTP(S) destinations are intended to be used.

Important limitations

• Dynamic websites can change their selectors and behavior over time.
• Some websites use bot protection that can block automated browsers.
• Login flows with CAPTCHA or complex multi-factor authentication may require a more specialized solution.
• waitForNavigation should only be enabled for clicks that are expected to trigger a real page navigation.
• The Actor is intended for public websites and does not provide access to private networks.

Responsible use

Only automate websites and data you are authorized to access. Follow applicable laws, website terms, privacy requirements, and rate limits.

This is an independent automation tool built with the open-source Puppeteer library. It is not affiliated with or endorsed by Google, Chrome, or the Puppeteer project.

Support

When reporting an issue, include:

• the public target URL
• a sanitized version of the workflow input
• the failing step number
• the error shown in the Dataset

Do not include passwords, API keys, private cookies, or other secrets in public issues.