Ghost CMS CVE-2026-105644

Exploitation requires the Ghost content import feature and an Administrator who is tricked into importing an attacker-supplied file containing a crafted SVG; the malicious SVG is then stored unsanitized and served from the site's own domain. Affected only in versions 4.0.0 up to (but not including) the fixed release. Key limiting factors: the attack is not self-service - it needs active social engineering of a privileged Administrator to perform the import (UI:R), and impact further depends on a staff user subsequently rendering the stored SVG in an authenticated admin context for session compromise. No special non-default configuration is otherwise required.

This is a genuine but conditional risk rather than a drop-everything priority. The supplied CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:N, 6.8 Medium) is internally consistent with the description: network-reachable but gated behind high attack complexity (AC:H) and required user interaction (UI:R) - specifically, an Administrator must be convinced to import a crafted file. No EPSS score was provided, there is no CISA KEV listing, and no public POC was supplied, so by the precision rules exploitation status is 'no public exploit identified at time of analysis.' The signals do not conflict: a Medium CVSS with AC:H/UI:R aligns with the absence of KEV/POC, indicating a stored-XSS issue that requires social engineering of a privileged user rather than mass, automatable exploitation. The main caveat is a version discrepancy in the source data (see confidence_notes) and the lack of EPSS/POC telemetry, which are missing rather than contradictory. Overall: patch on a normal cadence; it is not a wormable or unauthenticated-RCE emergency, but successful exploitation does yield high-value admin session compromise.

An attacker crafts an SVG file containing embedded JavaScript and, through phishing or a plausible pretext (e.g. a content migration or guest-post package), persuades a Ghost Administrator to import it via the content import feature. Because SVGs in imports were stored without sanitization (versions 4.0.0 through pre-6.67.0), the script is now hosted on the site's own domain and executes in the browser of staff users who view it, enabling theft or hijacking of their admin sessions. No public exploit code was identified at time of analysis, and exploitation hinges on convincing a privileged user to perform the import.