ghsa

### Impact This security advisory is a part of IBEXA-SA-2025-003, which resolves XSS vulnerabilities in several parts of the back office of Ibexa DXP. Back office access and varying levels of editing and management permissions are required to exploit these vulnerabilities. This typically means Editor or Administrator role, or similar. Injected XSS is persistent and can be reflected in the front office, possibly affecting end users. The fixes ensure XSS is escaped, and any existing injected XSS is rendered harmless. ### Patches - See "Patched versions". - https://github.com/ibexa/admin-ui/commit/72a64d90d249e5f4c4a5e8238f5d627c9b68d9b8 ### Workarounds None.

### Impact This security advisory is a part of IBEXA-SA-2025-003, which resolves XSS vulnerabilities in several parts of the back office of Ibexa DXP. Back office access and varying levels of editing and management permissions are required to exploit these vulnerabilities. This typically means Editor or Administrator role, or similar. Injected XSS is persistent and can be reflected in the front office, possibly affecting end users. The fixes ensure XSS is escaped, and any existing injected XSS is rendered harmless. ### Patches - See "Patched versions". - https://github.com/ibexa/admin-ui-assets/commit/219b71b70aaea9321947d2dbeb49fff1b49e05f4 ### Workarounds None.

### Impact This security advisory is a part of IBEXA-SA-2025-003, which resolves XSS vulnerabilities in several parts of the back office of Ibexa DXP. Back office access and varying levels of editing and management permissions are required to exploit these vulnerabilities. This typically means Editor or Administrator role, or similar. Injected XSS is persistent and can be reflected in the front office, possibly affecting end users. The fixes ensure XSS is escaped, and any existing injected XSS is rendered harmless. ### Patches - See "Patched versions". - https://github.com/ezsystems/ezplatform-admin-ui-assets/commit/219b71b70aaea9321947d2dbeb49fff1b49e05f4 ### Workarounds None.

### Impact This security advisory is a part of IBEXA-SA-2025-003, which resolves XSS vulnerabilities in several parts of the back office of Ibexa DXP. Back office access and varying levels of editing and management permissions are required to exploit these vulnerabilities. This typically means Editor or Administrator role, or similar. Injected XSS is persistent and can be reflected in the front office, possibly affecting end users. The fixes ensure XSS is escaped, and any existing injected XSS is rendered harmless. ### Patches - See "Patched versions". - https://github.com/ezsystems/ezplatform-admin-ui/commit/acaa620d4ef44e7c20908dc389d48064f2c19e6d ### Workarounds None.

### Summary Various date messages returned by `Language::userDate` are inserted into raw HTML, allowing anybody who can edit those messages to insert arbitrary HTML into the DOM. ### Details The result of `$this->lang->userDate( $timestamp, $this->user )` returns unescaped values, but is inserted as raw HTML by Citizen: https://github.com/StarCitizenTools/mediawiki-skins-Citizen/blob/072e4365e9084e4b153eac62d3666566c06f5a49/includes/Components/CitizenComponentUserInfo.php#L55-L60 ### PoC 1. Go to any page using citizen with the uselang parameter set to x-xss and while being logged in Depending on the registration date of the account you're logged in with, various messages can be shown. In my case, it's `november`:  ### Impact This impacts wikis where a group has the `editinterface` but not the `editsitejs` user right.

### Summary All system messages in menu headings using the Menu.mustache template are inserted as raw HTML, allowing anybody who can edit those messages to insert arbitrary HTML into the DOM. ### Details The system messages for menu headings are inserted unescaped into raw HTML: https://github.com/StarCitizenTools/mediawiki-skins-Citizen/blob/072e4365e9084e4b153eac62d3666566c06f5a49/templates/Menu.mustache#L8-L10 ### PoC 1. Go to any article using citizen with the `uselang` parameter set to `x-xss` 2. A large number of alerts will be shown for various messages, e.g.:   On the main page of my test wiki, the following messages were shown: `navigation`, `notifications`, `user-interface-preferences`, `personaltools`, `variants`, `views`, `associated-pages`, `cactions` and `toolbox`. ### Impact This impacts wiki...

### Summary Various preferences messages are inserted into raw HTML, allowing anybody who can edit those messages to insert arbitrary HTML into the DOM. ### Details The `innerHtml` of the label div is set to the `textContent` of the label, essentially unsanitizing the system messages: https://github.com/StarCitizenTools/mediawiki-skins-Citizen/blob/407052e7069bdeae927d6f1a2a1c9a45b473bf9a/resources/skins.citizen.preferences/addPortlet.polyfill.js#L18 ### PoC 1. Edit `citizen-feature-custom-font-size-name` (or any other message displayed in a heading in the preferences menu) to `<img src="" onerror="alert('citizen-feature-custom-font-size-name')">` (script tags don't work here due to the way the HTML is inserted) 2. Open the preferences menu 

### Summary The `citizen-search-noresults-title` and `citizen-search-noresults-desc` system messages are inserted into raw HTML, allowing anybody who can edit those messages to insert arbitrary HTML into the DOM. ### Details The system messages are inserted as raw HTML by the mustache template: https://github.com/StarCitizenTools/mediawiki-skins-Citizen/blob/407052e7069bdeae927d6f1a2a1c9a45b473bf9a/resources/skins.citizen.search/templates/TypeaheadPlaceholder.mustache#L8-L9 ### PoC 1. Edit `citizen-search-noresults-title` and `citizen-search-noresults-desc` to `<img src="" onerror="alert('citizen-search-noresults-title')">` and `<img src="" onerror="alert('citizen-search-noresults-desc')">` (script tags don't work here due to the way the HTML is inserted) 2. Open the search bar and search for a page that doesn't exist to get the "no results" messages to show up  ![image](https://github.com/user...

Worker process denial of service through file read operation. .A vulnerability exists in the Master's “pub_ret” method which is exposed to all minions. The un-sanitized input value “jid” is used to construct a path which is then opened for reading. An attacker could exploit this vulnerabilities by attempting to read from a filename that will not return any data, e.g. by targeting a pipe node on the proc file system.

Directory traversal vulnerability in recv_file method allows arbitrary files to be written to the master cache directory.