# SOME - Same Origin Method Execution
{% hint style="success" %}
Learn & practice AWS Hacking:[**HackTricks Training AWS Red Team Expert (ARTE)**](https://training.hacktricks.xyz/courses/arte)\
Learn & practice GCP Hacking: [**HackTricks Training GCP Red Team Expert (GRTE)**](https://training.hacktricks.xyz/courses/grte)
Support HackTricks
* Check the [**subscription plans**](https://github.com/sponsors/carlospolop)!
* **Join the** 💬 [**Discord group**](https://discord.gg/hRep4RUj7f) or the [**telegram group**](https://t.me/peass) or **follow** us on **Twitter** 🐦 [**@hacktricks\_live**](https://twitter.com/hacktricks\_live)**.**
* **Share hacking tricks by submitting PRs to the** [**HackTricks**](https://github.com/carlospolop/hacktricks) and [**HackTricks Cloud**](https://github.com/carlospolop/hacktricks-cloud) github repos.
{% endhint %}
### [WhiteIntel](https://whiteintel.io)
[**WhiteIntel**](https://whiteintel.io) is a **dark-web** fueled search engine that offers **free** functionalities to check if a company or its customers have been **compromised** by **stealer malwares**.
Their primary goal of WhiteIntel is to combat account takeovers and ransomware attacks resulting from information-stealing malware.
You can check their website and try their engine for **free** at:
{% embed url="https://whiteintel.io" %}
***
## Same Origin Method Execution
There will be occasions where you can execute some limited javascript in a page. For example, in the case where you can[ **control a callback value that will be executed**](./#javascript-function).
In those case, one of the best things that you could do is to **access the DOM to call whatever** sensitive action you can find in there (like clicking a button). However, usually you will find this vulnerability in **small endpoints without any interesting thing in the DOM**.
In those scenarios, this attack will be very useful, because its goal is to be able to **abuse the limited JS execution inside a DOM from a different page from the same domain** with much interesting actions.
Basically, the attack flow is the following:
* Find a **callback that you can abuse** (potentially limited to \[\w\\.\_]).
* If it's not limited and you can execute any JS, you could just abuse this as a regular XSS
* Make the **victim open a page** controlled by the **attacker**
* The **page will open itself** in a **different window** (the new window will have the object **`opener`** referencing the initial one)
* The **initial page** will load the **page** where the **interesting DOM** is located.
* The **second page** will load the **vulnerable page abusing the callback** and using the **`opener`** object to **access and execute some action in the initial page** (which now contains the interesting DOM).
{% hint style="danger" %}
Note that even if the initial page access to a new URL after having created the second page, the **`opener` object of the second page is still a valid reference to the first page in the new DOM**.
Moreover, in order for the second page to be able to use the opener object **both pages must be in the same origin**. This is the reason why, in order to abuse this vulnerability, you need to find some sort of **XSS in the same origin**.
{% endhint %}
### Exploitation
* You can use this form to **generate a PoC** to exploit this type of vulnerability: [https://www.someattack.com/Playground/SOMEGenerator](https://www.someattack.com/Playground/SOMEGenerator)
* In order to find a DOM path to a HTML element with a click you can use this browser extension: [https://www.someattack.com/Playground/targeting\_tool](https://www.someattack.com/Playground/targeting\_tool)
### Example
* You can find a vulnerable example in [https://www.someattack.com/Playground/](https://www.someattack.com/Playground/)
* Note that in this example the server is **generating javascript code** and **adding** it to the HTML based on the **content of the callback parameter:** `` . Thats why in this example you don't need to indicate the use of `opener` explicitly.
* Also check this CTF writeup: [https://ctftime.org/writeup/36068](https://ctftime.org/writeup/36068)
## References
* [https://conference.hitb.org/hitbsecconf2017ams/sessions/everybody-wants-some-advance-same-origin-method-execution/](https://conference.hitb.org/hitbsecconf2017ams/sessions/everybody-wants-some-advance-same-origin-method-execution/)
{% hint style="success" %}
Learn & practice AWS Hacking:[**HackTricks Training AWS Red Team Expert (ARTE)**](https://training.hacktricks.xyz/courses/arte)\
Learn & practice GCP Hacking: [**HackTricks Training GCP Red Team Expert (GRTE)**](https://training.hacktricks.xyz/courses/grte)
Support HackTricks
* Check the [**subscription plans**](https://github.com/sponsors/carlospolop)!
* **Join the** 💬 [**Discord group**](https://discord.gg/hRep4RUj7f) or the [**telegram group**](https://t.me/peass) or **follow** us on **Twitter** 🐦 [**@hacktricks\_live**](https://twitter.com/hacktricks\_live)**.**
* **Share hacking tricks by submitting PRs to the** [**HackTricks**](https://github.com/carlospolop/hacktricks) and [**HackTricks Cloud**](https://github.com/carlospolop/hacktricks-cloud) github repos.
{% endhint %}