Package Instance

Flarum notifications can leak restricted content
Using the notifications feature, one can read restricted/private content and bypass access checks that would be in place for such content.

The notification-sending component does not check that the subject of the notification can be seen by the receiver, and proceeds to send notifications through their different channels. The alerts do not leak data despite this as they are listed based on a visibility check, however, emails are still sent out.

This means that, for extensions which restrict access to posts, any actor can bypass the restriction by subscribing to the discussion if the [*Subscriptions*](https://extiverse.com/extension/flarum/subscriptions) extension is enabled.

### Impact
The attack allows the leaking of some posts in the forum database, including posts awaiting approval, posts in tags the user has no access to if they could subscribe to a discussion before it becomes private, and posts restricted by third-party extensions.

Other leaks could also happen for different notification subjects if some features allowed to receive specific types of notifications for restricted content.

All Flarum versions prior to v1.6.3 are affected.

### Patches
The vulnerability has been fixed and published as flarum/core v1.6.3. All communities running Flarum should upgrade as soon as possible to v1.6.3 using:

```
composer update --prefer-dist --no-dev -a -W
```
You can then confirm you run the latest version using:

```
composer show flarum/core
```

### Workarounds
Disable the Flarum Subscriptions extension or disable email notifications altogether.

**There is no other supported workaround for this issue for Flarum versions below 1.6.3.**

### For more information
For any questions or comments on this vulnerability please visit https://discuss.flarum.org/

For support questions create a discussion at https://discuss.flarum.org/t/support.

A reminder that if you ever become aware of a security issue in Flarum, please report it to us privately by emailing [security@flarum.org](mailto:security@flarum.org), and we will address it promptly.

Flarum's logout Route allows open redirects
The Flarum `/logout` route includes a redirect parameter that allows any third party to redirect users from a (trusted) domain of the Flarum installation to redirect to any link. Sample: `example.com/logout?return=https://google.com`. For logged-in users, the logout must be confirmed. Guests are immediately redirected. This could be used by spammers to redirect to a web address using a trusted domain of a running Flarum installation.

Some ecosystem extensions modifying the logout route have already been affected. Sample: https://discuss.flarum.org/d/22229-premium-wordpress-integration/526

Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')
flarum is a forum software package for building communities. In versions prior to 1.7.0 an admin account which has already been compromised by an attacker may use a vulnerability in the `LESS` parser which can be exploited to read sensitive files on the server through the use of path traversal techniques. An attacker can achieve this by providing an absolute path to a sensitive file in the custom `LESS` setting, which the `LESS` parser will then read. For example, an attacker could use the following code to read the contents of the `/etc/passwd` file on a linux machine. The scope of what files is vulnerable will depend on the permissions given to the running flarum process. The vulnerability has been addressed in version `1.7`. Users should upgrade to this version to mitigate the vulnerability. Users unable to upgrade may mitigate the vulnerability by ensuring that their admin accounts are secured with strong passwords and follow other best practices for account security. Additionally, users can limit the exposure of sensitive files on the server by implementing appropriate file permissions and access controls at the operating system level.

Flarum: Path traversal in LESS parser via theme color settings (incomplete fix for CVE-2023-27577)
## Summary

Flarum's patch for [CVE-2023-27577](https://github.com/flarum/framework/security/advisories/GHSA-vhm8-wwrf-3gcw) restricted the `@import` and `data-uri()` LESS features in the `custom_less` setting, but the same restriction was never applied to other settings registered as LESS config variables (for example `theme_primary_color` and `theme_secondary_color`, as well as any key registered via `Extend\Settings::registerLessConfigVar()`).

Those values are interpolated verbatim into the LESS source at compile time, allowing an authenticated administrator to craft a theme-color value that injects an arbitrary `@import` directive into the compiled `forum.css`. Because the underlying LESS parser honours `@import (inline) '<path>'`, an attacker can read arbitrary files reachable by the PHP process (local file inclusion) or trigger outbound HTTP(S) requests (server-side request forgery).

## Impact

An attacker who has compromised — or legitimately obtained — an administrator account can:

- **Read arbitrary local files** reachable by the PHP process (e.g. `/etc/passwd`, `.env`, config files containing database credentials, OAuth secrets, API keys).
- **Trigger outbound HTTP/HTTPS requests** from the Flarum host, enabling SSRF against internal services and cloud metadata endpoints such as `http://169.254.169.254/` (AWS IMDSv1, GCP, Azure).

The contents of the attacker-controlled import are embedded into the compiled `forum.css`, which is publicly served — so the attacker can retrieve whatever was read simply by fetching the CSS file.

This is a privilege-escalation vulnerability: a forum administrator is not intended to have host-level file read or access to internal network resources.

### Example payload

Submitted via `POST /api/settings` with an admin session:

```json
{ "theme_primary_color": "#4D698E;@import (inline) '/etc/passwd';" }
```

The setting is stored verbatim, interpolated into the LESS source on the next CSS compile, and the target file's contents appear in `/assets/forum.css`.

## Patches

- **`flarum/core` 1.8.16** — fix for the 1.x branch.
- **`flarum/core` 2.0.0-rc.1** — fix for the 2.x branch.

The fix extends the existing `@import` / `data-uri()` validation in `Flarum\Forum\ValidateCustomLess::whenSettingsSaving` to every dirty setting whose key is registered as a LESS config variable, not just `custom_less`.

## Workarounds

If upgrading is not immediately possible:

- Ensure administrator accounts are protected with strong, unique passwords and (where supported) two-factor authentication.
- Restrict administrator access to trusted users only.
- Review the forum's public `forum.css` for unexpected content that could indicate prior exploitation.

There is no configuration-level mitigation on affected versions — the fix requires the upgraded code.

## Resources

- [CVE-2023-27577](https://nvd.nist.gov/vuln/detail/CVE-2023-27577) — the original vulnerability whose patch was incomplete.
- [GHSA-vhm8-wwrf-3gcw](https://github.com/flarum/framework/security/advisories/GHSA-vhm8-wwrf-3gcw) — the original advisory.

## Credit

Reported to the Flarum Foundation by **William (Liam) Snow IV** ([@LiamSnow](https://github.com/LiamSnow)), discovered during a graduate-level network security lab at Worcester Polytechnic Institute.

Flarum post mentions can be used to read any post on the forum without access control
Using the mentions feature provided by the flarum/mentions extension, users can mention any post ID on the forum with the special `@"<username>"#p<id>` syntax.

The following behavior never changes no matter if the actor should be able to read the mentioned post or not:

A URL to the mentioned post is inserted into the actor post HTML, leaking its discussion ID and post number.

The `mentionsPosts` relationship included in the `POST /api/posts` and `PATCH /api/posts/<id>` JSON responses leaks the full JSON:API payload of all mentioned posts without any access control. This includes the content, date, number and attributes added by other extensions.

An attacker only needs the ability to create new posts on the forum to exploit the vulnerability. This works even if new posts require approval. If they have the ability to edit posts, the attack can be performed even more discreetly by using a single post to scan any size of database and hiding the attack post content afterward.

### Impact
The attack allows the leaking of all posts in the forum database, including posts awaiting approval, posts in tags the user has no access to, and private discussions created by other extensions like FriendsOfFlarum Byobu. This also includes non-comment posts like tag changes or renaming events. 

The discussion payload is not leaked but using the mention HTML payload it's possible to extract the discussion ID of all posts and combine all posts back together into their original discussions even if the discussion title remains unknown.

All Flarum versions prior to `v1.6.3` are affected.

### Patches
The vulnerability has been fixed and published as flarum/core v1.6.3. All communities running Flarum have to upgrade as soon as possible to v1.6.3 using:

```
composer update --prefer-dist --no-dev -a -W
```
You can then confirm you run the latest version using:

```
composer show flarum/core
```

### Workarounds
Disable the mentions extension.

### For more information
For any questions or comments on this vulnerability please visit https://discuss.flarum.org/

For support questions create a discussion at https://discuss.flarum.org/t/support.

A reminder that if you ever become aware of a security issue in Flarum, please report it to us privately by emailing [security@flarum.org](mailto:security@flarum.org), and we will address it promptly.

Flarum Vulnerable to Session Hijacking via Authoritative Subdomain Cookie Overwrite
A session hijacking vulnerability exists when an attacker-controlled **authoritative subdomain** under a parent domain (e.g., `subdomain.host.com`) sets cookies scoped to the parent domain (`.host.com`). This allows session token replacement for applications hosted on sibling subdomains (e.g., `community.host.com`) if session tokens aren't rotated post-authentication.

**Key Constraints**:
- Attacker must control **any subdomain** under the parent domain (e.g., `evil.host.com` or `x.y.host.com`).
- Parent domain must **not** be on the [Public Suffix List](https://publicsuffix.org/).

Due to non-existent session token rotation after authenticating we can theoretically reproduce the vulnerability by using browser dev tools, but due to the browser's security measures this does not seem to be exploitable as described.

---