CVE-2025-10854 Symlink Following in txtai leads to arbitrary file write when loading untrusted embedding indices
Attackers can write arbitrary files to the filesystem by loading malicious embedding indices containing symbolic links in the txtai library.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Over the internet
- Victim action needed
- Not needed
- Fix
- None listed
What is CVE-2025-10854?
CVE-2025-10854 is a high-severity path traversal vulnerability, published on 22 Sep 2025. An attacker can exploit it over the internet without logging in, with no action from the victim.
Official description
The txtai framework allows the loading of compressed tar files as embedding indices. While the validate function is intended to prevent path traversal vulnerabilities by ensuring safe filenames, it does not account for symbolic links within the tar file. An attacker is able to write a file anywhere in the filesystem when txtai is used to load untrusted embedding indices
Weakness: CWE-61
Which versions are affected by CVE-2025-10854?
How do I fix CVE-2025-10854?
No fix has been published for CVE-2025-10854 yet. Limit exposure of the affected software until one is available.
Upgrade, and resolve every user-supplied path against an allowed base folder before opening it. More about Path traversal.
Is CVE-2025-10854 being exploited?
There is no known exploitation of CVE-2025-10854 in the wild and no public exploit code.
Sources
Also known as GHSA-q235-c2j2-rm76.