Tasks

Virtual Worker Send Commands

Executes a set of virtual worker commands to perform automated actions

See virtual-worker concept for more details.

# Basic usage

Run a virtual worker that navigates to a website and types text into a field.

# Parameters

# timeout_seconds

optional

number

The seconds after which the virtual_worker_send_commands task will time out.
This is also used as the timeout for starting a new worker inside a cluster, specifically timeout_seconds - 10, to allow for actual command execution before the overall task timeout.

Default: 300

# runtime

Object

required

Defines the execution settings for the virtual worker.

# commands

Object

required

A set of commands to execute by the virtual worker.
Each command is defined as a key-value pair, where the key is a unique identifier for the command, and the value is an object defining the command type and its parameters. See more below.

IMPORTANT :
Each command must have a unique __order parameter to define the execution order.

# Browser Commands

# slow_mo

optional

number

Every browser command below accepts an optional <command>.params.slow_mo parameter that defines a delay in milliseconds between each action performed by the browser.
Useful for debugging or visualizing the actions being performed.

Needs to be set on the first command to take effect for the entire browser session.

# Browser Write text

Will write the specified text into the element located by the locator.

# Browser Navigate

Will navigate the browser to the specified URL.

# Browser Click

Will click on the element located by the locator.

# Browser Press Key

Will press the specified key in the focused element.

# Browser Get Local Storage

Will retrieve the entire local storage of the browser. Depends on the website what data is stored there.

# Browser Extract Data

Will extract data from the page using the specified locators.
Converts HTML elements into structured JSON data.

Specifying a string as the value of a locator will extract the text content of the element.

Specifying an object as the value of a locator allows for more complex extraction, including lists and attributes.

To extract a list of elements, use the special key _list with a locator string that will match multiple elements. To extract properties of each list item, use the special key _list_item_properties with an object defining the properties to extract.

To extract an attribute of an element, use the special keys _locator and _target within a locator object.

# Browser Screenshot

Will take a screenshot of the current page or of the given locator.

params.locator (optional): A locator string to capture a specific element's screenshot.

params.screenshot_options (optional) parameter can be used to specify additional options for the screenshot.
Supports all options from Playwright's page.screenshot() method.

Notably:

• fullPage (boolean): When true, takes a screenshot of the full scrollable page. Default is false.
• clip (object): An object specifying the clipping region of the screenshot with properties x, y, width, and height.

Returns

# Browser Extract HTML

Will extract the HTML content of the current page or of the given locator.

params.locator (optional): A locator string to extract HTML from a specific element.

Returns

# Locators

Locators are used to identify and interact with elements on a web page. They can be defined using various strategies, such as CSS selectors, XPath expressions, or text content. See locators documentation for more details and examples.

# Browser Session Management

params.session_id (string, optional) parameter can be provided to maintain a persistent browser session across multiple commands.

If provided, the virtual worker will maintain a persistent browser session associated with that session ID.

Default: {{stub.stubref}}

# File Server Commands

# Download Files

Integrates with the Stubber file storage system to download files by their UUIDs.
Specify either an array of strings (fileuuids) or an array of objects with a fileuuid property.

Files are downloaded to the default tmp directory on the virtual worker machine.

# Upload Files

Uploads files from the virtual worker machine to the Stubber file storage system.
Specify an array of file paths to be uploaded.
Resulting fileuuids are available in the command payload if successful.

# CLI Command

You can also execute arbitrary CLI commands on the virtual worker.

# Run CLI Command

params.command (string, required): The CLI command to execute on the virtual worker.

Returns

        {
  "run_cli_command": {
    "success": true,
    "payload": {
      "stdout": "Hello, World!",
      "stderr": "",
    }
  }
}

      

# Available Libraries

The following libraries are pre-installed and can be used in CLI commands:

• curl
• unzip
• tree

# Python Libraries

The following Python libraries are pre-installed and can be used in python scripts:

• pandas
• openpyxl
• odfpy
• numpy
• xlsxwriter

To use python, simply write a script and execute it via the cli_run command. For example:

        echo "print('Hello from Python!')" > /tmp/script.py && python /tmp/script.py