DOM-based Cross-Site Scripting (XSS) in OpenEMR 7.0.0 and below at White list files in openemr/openemr
Reported on
Jul 22nd 2022
Description
We would like to report the vulnerability we found during software testing. The OpenEMR 7.0.0 (latest version) and below version; Open Source electronic health records and medical practice management application; has DOM-based Cross-Site Scripting (XSS) vulnerability in the add-manually-input field on white list file page that never been reported before (We've checked from CVE Official website).
Vulnerability Type
DOM-based Cross-Site Scripting (XSS)
Affected Page/URL
https://<openemrurl>/interface/super/manage_site_files.php
Sample XSS Payload
"><script>alert(`CVE_Hunting`)</script>
Vulnerable Source Code
/var/www/localhost/htdocs/openemr/interface/super/manage_site_files.php (Please see more details in the occurrences section)
Implication
Client-side scripts are used extensively by modern web applications. They perform from simple functions (such as the formatting of text) up to full manipulation of client-side data and Operating System interaction.
Unlike traditional Cross-Site Scripting (XSS), where the client is able to inject scripts into a request and have the server return the script to the client, DOM XSS does not require that a request be sent to the server and may be abused entirely within the loaded page.
This occurs when elements of the DOM (known as the sources) are able to be manipulated to contain untrusted data, which the client-side scripts (known as the sinks) use or execute an unsafe way.
Scanner has discovered that by inserting an HTML element into the page's DOM inputs (sources), it was possible to then have the HTML element rendered as part of the page by the sink.
Recommendation
Client-side document rewriting, redirection, or other sensitive action, using untrusted data, should be avoided wherever possible, as these may not be inspected by server side filtering.
To remedy DOM XSS vulnerabilities where these sensitive document actions must be used, it is essential to
- Ensure any untrusted data is treated as text, as opposed to being interpreted as code or mark-up within the page.
- Escape untrusted data prior to being used within the page. Escaping methods will vary depending on where the untrusted data is being used. (See references for details.)
- Use
document.createElement
,element.setAttribute
,element.appendChild
, etc. to build dynamic interfaces as opposed to HTML rendering methods such asdocument.write
,document.writeIn
,element.innerHTML
, orelement.outerHTML
etc.
Discoverer/Reporters
- Ammarit Thongthua, Rattapon Jitprajong and Nattakit Intarasorn from Secure D Center Research Team
Example PoC Screenshots
OpenEMR Version 6.1.0
OpenEMR Version 7.0.0
Admin > System > Files
At the White list we Injected malicious JavaScript Payload in to {add-manually-input} with Sample XSS Payload
"><script>alert(`CVE_Hunting`)</script>
Renderred Malicious JavaScript (XSS)
Impact
DOM-based vulnerabilities arise when a website contains JavaScript that takes an attacker-controllable value, known as a source, and passes it into a dangerous function, known as a sink.
Occurrences
manage_site_files.php L537-L541
$('#add-manually').on('click', function () {
var new_type = $("#add-manually-input").val();
if(new_type.length < 1)return;
$('#white-list').prepend("<option value="+new_type+">"+new_type+"</option>")
})
Thanks for the report. We are working on a fix.
A preliminary fix has been posted in commit 74d21039aec641b2c406e3baf238ae4602a968b6
Please do not create a CVE # or make this vulnerability public at this time. I will make this fix official about 1 week after we release 7.0.0 patch 1 (7.0.0.1), which will likely be in about 3-7 weeks. After I do that, then will be ok to make CVE # and make it public.
Thanks!
Just waiting on the go-ahead from the maintainer, then I can arrange and publish the CVE π
OpenEMR patch 1 (7.0.0.1) has been released, so this has been fixed. You have permission to make CVE # and make this public.
Dear @Admin, @Jamie Slome Could you please help to assign the CVE for this finding please? Thanks and appreciate for your help π
@Brady Miller Thanks for your response π
CVE assigned and will be automatically published in the next couple of hours π β₯οΈ
@Jamie Slome Appreciate for your contribution πβ₯οΈ