Описание
jquery.terminal self XSS on user input
Impact
This is low impact and limited XSS, because code for XSS payload is always visible, but attacker can use other techniques to hide the code the victim sees.
Also if the application use execHash option and execute code from URL the attacker can use this URL to execute his code. The scope is limited because the javascript code inside html attribute used is added to span tag, so no automatic execution like with onerror on images is possible.
Patches
Fixed version 2.31.1
Workarounds
The user can use formatting that wrap whole user input and it's no op.
The fix will only work when user of the library is not using different formatters (e.g. to highlight code in different way).
References
The issue was reported here jcubic/jquery.terminal#727
For more information
If you have any questions or comments about this advisory:
- Open an issue in jcubic/jquery.terminal
- Email us at jcubic@onet.pl
Ссылки
- https://github.com/jcubic/jquery.terminal/security/advisories/GHSA-x9r5-jxvq-4387
- https://nvd.nist.gov/vuln/detail/CVE-2021-43862
- https://github.com/jcubic/jquery.terminal/issues/727
- https://github.com/jcubic/jquery.terminal/commit/77eb044d0896e990d48a9157f0bc6648f81a84b5
- https://github.com/jcubic/jquery.terminal
- https://github.com/jcubic/jquery.terminal/releases/tag/2.31.1
Пакеты
jquery.terminal
< 2.31.1
2.31.1
Связанные уязвимости
jQuery Terminal Emulator is a plugin for creating command line interpreters in your applications. Versions prior to 2.31.1 contain a low impact and limited cross-site scripting (XSS) vulnerability. The code for XSS payload is always visible, but an attacker can use other techniques to hide the code the victim sees. If the application uses the `execHash` option and executes code from URL, the attacker can use this URL to execute their code. The scope is limited because the javascript attribute used is added to span tag, so no automatic execution like with `onerror` on images is possible. This issue is fixed in version 2.31.1. As a workaround, the user can use formatting that wrap whole user input and its no op. The code for this workaround is available in the GitHub Security Advisory. The fix will only work when user of the library is not using different formatters (e.g. to highlight code in different way).