~ / search / GHSA-67c6-q4j4-hccg

GHSA-67c6-q4j4-hccg

Flarum vulnerable to LFI and Blind SSRF via Avatar upload

Source: GitHub Advisory Database (reviewed) High CVE-2023-40033

Packages

EcosystemPackageAffectedFixed in
Packagist flarum/core < 1.8.0 1.8.0
Packagist flarum/framework < 1.8.0 1.8.0

Details

As written in the advisory (Markdown source, shown as plain text).

## Impact The Flarum forum software is affected by a vulnerability that allows an attacker to conduct a Blind SSRF attack or disclose any file on the server, even with a basic user account on any Flarum forum. By uploading a file containing a URL and spoofing the MIME type, an attacker can manipulate the application to execute unintended actions. The vulnerability is due to the behavior of the `intervention/image` package, which attempts to interpret the supplied file contents as a URL, which then fetches its contents. This allows an attacker to exploit the vulnerability to perform SSRF attacks, disclose local file contents, or conduct a blind oracle attack. ### Patches This has been patched in Flarum **v1.8**. ## Workarounds As a temporary workaround for the SSRF aspect of the vulnerability, one can disable PHP's `allow_url_fopen` which will prevent the fetching of external files via URLs. ### Credits Adam Kues - [Assetnote](https://assetnote.io/)

References

External links from the advisory. They lead to third-party sites.