Data Extraction Actions

Actions for reading data from page elements.

getText

Get the text content of an element, without any HTML tags.

Value: string (selector)

<h1 class="title">Welcome to BrowserWorker</h1>

[
	{ "getText": "h1.title" }
]

Returns: "Welcome to BrowserWorker"

Save for later use:

[
	{ "getText": ".product-price", "options": { "save": "price" } },
	{ "fill": ["input#discount", "${price}"] }
]

Append to an array:

[
	{ "getText": ".item:nth-child(1)", "options": { "push": "names" } },
	{ "getText": ".item:nth-child(2)", "options": { "push": "names" } }
	// Get result by getSavedData
	{ "getSavedData": "names" }
]

Result of getSavedData:

{
	"names": ["Item 1", "Item 2"]
}

Options

Option	Type	Default	Description
save	string	—	Save the result with this key for use with ${key}
push	string	—	Append the result to an array in the data cache
waitFor	boolean	true	Wait for the element to appear
timeout	number	10000	How long to wait (ms)
iframe	string | string[]	—	Look inside an iframe

getHTML

Get the inner HTML of an element. By default it returns only the contents — use outer: true to get the element itself including its own tag.

Value: string (selector)

<div class="content"><p>Hello <strong>World</strong></p></div>

[
	{ "getHTML": ".content" }
]

Returns: "<p>Hello <strong>World</strong></p>"

Include the element's own tag:

[
	{
		"getHTML": "#article",
		"options": { "outer": true, "save": "articleHTML" }
	}
]

Capture the full rendered page (current DOM state after scripts run) by targeting the root html element with outer: true:

{
	"actions": [
		{ "open": "https://example.com" },
		{ "getHTML": "html", "options": { "outer": true } }
	]
}

Options

Option	Type	Default	Description
outer	boolean	false	Return the full element HTML instead of just its contents
save	string	—	Save the result
push	string	—	Append to an array
waitFor	boolean	true	Wait for the element to appear
timeout	number	10000	How long to wait (ms)
iframe	string | string[]	—	Look inside an iframe

getAttribute

Get the value of an attribute from an element.

Value: [selector, attribute]

<a class="main-link" href="https://example.com">Visit</a>

[
	{ "getAttribute": ["a.main-link", "href"] }
]

Returns: "https://example.com"

Save the result:

[
	{ "getAttribute": ["img.avatar", "src"], "options": { "save": "avatarUrl" } }
]

Options

Option	Type	Default	Description
save	string	—	Save the result
push	string	—	Append to an array
waitFor	boolean	true	Wait for the element to appear
timeout	number	10000	How long to wait (ms)
iframe	string | string[]	—	Look inside an iframe

getValue

Get the current value of a form field (<input>, <textarea>, or <select>).

Value: string (selector)

<input type="email" name="email" value="hello@example.com" />

[
	{ "getValue": "input[name=email]" }
]

Returns: "hello@example.com"

Save the result:

[
	{ "getValue": "select#country", "options": { "save": "selectedCountry" } }
]

Options

Option	Type	Default	Description
save	string	—	Save the result
push	string	—	Append to an array
waitFor	boolean	true	Wait for the element to appear
timeout	number	10000	How long to wait (ms)
iframe	string | string[]	—	Look inside an iframe

extractAll

Extract data from multiple matching elements at once. Each matched element becomes one item in the result.

Value: [containerSelector, fieldMap]

<div class="product-card">
	<h3>Laptop</h3>
	<span class="price">$999</span>
	<a href="/laptop">Details</a>
	<img src="/img/laptop.jpg" />
</div>
<div class="product-card">
	<h3>Phone</h3>
	<span class="price">$699</span>
	<a href="/phone">Details</a>
	<img src="/img/phone.jpg" />
</div>

[
	{
		"extractAll": [
			".product-card",
			{
				"title": "h3",
				"price": ".price",
				"link": ["a", "href"],
				"image": ["img", "src"]
			}
		]
	}
]

Returns:

[
	{ "title": "Laptop", "price": "$999", "link": "/laptop", "image": "/img/laptop.jpg" },
	{ "title": "Phone", "price": "$699", "link": "/phone", "image": "/img/phone.jpg" }
]

Field Map Syntax

The second argument defines what to extract. Keys are your field names, values are selectors or [selector, attribute] pairs:

Format	Description	Example
"selector"	Text content of a child element	"title": "h3"
["selector", "attr"]	Attribute value of a child element	"link": ["a", "href"]
":scope > ..."	Select relative to the container element	"desc": ":scope > p"
[":scope", "attr"]	Attribute on the container element itself	"id": [":scope", "data-id"]

Using :scope in selectors

Prefix a selector with :scope to match relative to each container, not the whole page. Useful when child elements share the same tag across multiple containers.

<div class="card">
	<p>Short intro</p>
	<div>
		<div>Posted on Jan 1</div>
		<div>Read the full article here.</div>
	</div>
</div>

[
	{
		"extractAll": [
			".card",
			{
				"intro": "p",
				"description": ":scope > div > div:nth-child(2)"
			}
		]
	}
]

Returns:

[
	{ "intro": "Short intro", "description": "Read the full article here." }
]

Using :scope to read container attributes

Use [":scope", "attr"] when the data you need is an attribute on the container element itself, not a child:

<div class="item" data-id="101">
	<h3>Laptop</h3>
	<span class="price">$999</span>
</div>
<div class="item" data-id="102">
	<h3>Phone</h3>
	<span class="price">$699</span>
</div>

[
	{
		"extractAll": [
			".item",
			{
				"id": [":scope", "data-id"],
				"title": "h3",
				"price": ".price"
			}
		]
	}
]

Returns:

[
	{ "id": "101", "title": "Laptop", "price": "$999" },
	{ "id": "102", "title": "Phone", "price": "$699" }
]

Extract from a table, limit to 10 rows:

[
	{
		"extractAll": [
			"table tbody tr",
			{
				"name": "td:nth-child(1)",
				"email": "td:nth-child(2)",
				"role": "td:nth-child(3)"
			}
		],
		"options": { "limit": 10, "save": "users" }
	}
]

Options

Option	Type	Default	Description
limit	number	0	Max items to return (0 = no limit)
save	string	—	Save the result array
push	string	—	Append to an array
waitFor	boolean	true	Wait for the container elements to appear
timeout	number	10000	How long to wait (ms)

getSavedData

Get values you saved earlier in the task. Each key maps to its saved value, or null if it was never set.

Value: string | string[]

Single key:

[
	{ "getSavedData": "price" }
]

Multiple keys:

[
	{ "getSavedData": ["price", "title", "users"] }
]

Returns:

{
	"price": "$19.99",
	"title": "Laptop",
	"users": null
}

clearSavedData

Delete all saved data. Use this when you want to start fresh mid-task.

Value: true

[
	{ "clearSavedData": true }
]

The data cache is cleared automatically in these situations:

• New task starts — cleared at the beginning of every task
• Chrome restarts — cleared on every browser launch

Data persists across session tasks

When tasks share the same session.id, the data cache is not cleared between them. This lets you save data in one API call and read it in the next. Use clearSavedData if you want to reset it between session tasks.

See Sessions for details.

---

Example: Scrape a Product Listing

[
	{ "openNewTab": "https://example.com/products" },
	{ "waitForElement": ".product-card" },
	{
		"extractAll": [
			".product-card",
			{
				"name": "h3.product-name",
				"price": ".price",
				"rating": ".stars",
				"url": ["a", "href"],
				"image": ["img", "src"]
			}
		],
		"options": { "limit": 20, "save": "products" }
	},
	{ "getSavedData": ["products"] }
]

Example: Collect Data from a Page

[
	{ "openNewTab": "https://example.com/page/1" },
	{ "getText": "h1", "options": { "save": "pageTitle" } },
	{ "getText": ".result-count", "options": { "save": "totalResults" } },
	{ "getAttribute": ["a.next-page", "href"], "options": { "save": "nextPageUrl" } },
	{ "getSavedData": ["pageTitle", "totalResults", "nextPageUrl"] }
]