Cross-Site Scripting in hMailServer Webmail Decryption

Cross-site scripting in the webmail of Progressive Robot hMailServer 6.3.2 through 6.3.5 allows a remote attacker who can send a user an encrypted message to run script in the webmail's origin with that user's session. When the webmail decrypted an S/MIME message (from 6.3.2) or an OpenPGP message (from 6.3.4) in the browser, it offered each decrypted attachment as a blob URL of the media type the message declared for it. A click saved the file, but if the user opened the attachment in a new tab, a part declared as text/html was rendered as a document of the webmail's origin and its script could read the mailbox, send mail and change the account through the REST API. The webmail is served only when the REST API is enabled, which it is not by default.

Casky was already ahead

This CVE exploits attack patterns that Casky's 323matched skills already investigate — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.

Analysis

This vulnerability affects Progressive Robot hMailServer versions 6.3.2 through 6.3.5, where a cross-site scripting (CWE-79) flaw exists in the webmail's handling of encrypted messages. When users decrypt S/MIME or OpenPGP messages in their browser, the application unsafely renders decrypted attachments as blob URLs with their declared media types. An attacker who can send an encrypted message to a target user can exploit this by crafting a malicious attachment that, when decrypted and opened, executes arbitrary JavaScript within the webmail's security context. This grants the attacker full access to the user's session, emails, and potentially sensitive data. The vulnerability affects any organization using vulnerable hMailServer versions with webmail functionality enabled, particularly those handling encrypted communications.

Casky's 323 mapped security skills identify this attack through detection of Initial Access (TA0001) and Execution (TA0043) techniques. Practitioners using Casky would observe findings related to: (1) suspicious blob URL generation patterns in webmail DOM manipulation, (2) unsafe handling of decrypted content without proper sanitization, (3) attachment rendering that trusts user-supplied media type declarations, and (4) script execution triggered by user interaction with decrypted attachments. Claude's extended reasoning capability would correlate these indicators across webmail logs, browser security events, and email metadata to surface the complete attack chain—from malicious message delivery through session compromise—enabling practitioners to identify exploitation attempts and prioritize patching or implementing additional input validation controls.

Live Threat IntelligenceComing soon

Casky Risk Score

—

EPSS Probability

—

Shodan Exposure

—

GreyNoise Signal

—

Composite risk scoring from EPSS, CISA KEV, Shodan, and GreyNoise — 21 security APIs correlated into a single Casky Risk Score. Coming in Casky Pro. Join early access →

323 Casky skills cover this attack pattern

These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-103647.

Abuse of Public-Facing API: Mobile API Abuse

mobile security · low

Run

abusing-dpapi-for-credential-access

red teaming · high

Run

abusing-shadow-credentials-for-privesc

red teaming · high

Run

MITRE ATT&CK Techniques

TA0001 TA0043

Investigate the attack patterns behind CVE-2026-103647

Casky has 323 skills that investigate the attack patterns behind CVE-2026-103647. Run one and get CVSS-scored findings in 3 minutes.

Run the skill that detects this →

X Instagram LinkedIn