Описание
Mailgen has HTML Injection and XSS Filter Bypass in Plaintext Emails
Summary
An HTML injection vulnerability in plaintext emails generated by Mailgen has been discovered. Projecta are affected if the Mailgen.generatePlaintext(email) method is used and passed in user-generated content. The issue was discovered and reported by Edoardo Ottavianelli (@edoardottt).
Details
The following function (inside index.js) is intended to strip all HTML content to produce a plaintext string.
The process fails because it searches for HTML tags and attempts to strip them from the input. However, if the unicode encoded characters are present inside HTML tags, they are not removed. These encoded tags are then decoded later and become valid HTML content, which can lead to XSS vulnerabilities.
A valid payload is: <img src=x onerror=alert(1)\u2028>.
PoC
Resulting output file (emailText.html):
Impact
Depending on the context/environment where the plaintext message is used, if HTML is rendered and executed can result in arbitrary code execution in the browser of the victim (potentially stealing secrets or sensitive information contained in the message).
Credits
Edoardo Ottavianelli (@edoardottt)
Пакеты
mailgen
<= 2.0.31
2.0.32
Связанные уязвимости
mailgen is a Node.js package that generates responsive HTML e-mails for sending transactional mail. Mailgen versions through 2.0.31 contain an HTML injection vulnerability in plaintext emails generated with the generatePlaintext method when user generated content is supplied. The plaintext generation code attempts to strip HTML tags using a regular expression and then decodes HTML entities, but tags that include certain Unicode line separator characters are not matched and removed. These encoded tags are later decoded into valid HTML content, allowing unexpected HTML to remain in output intended to be plaintext. Projects are affected if they call Mailgen.generatePlaintext with untrusted input and then render or otherwise process the returned string in a context where HTML is interpreted. This can lead to execution of attacker supplied script in the victim’s browser. Version 2.0.32 fixes the issue.