Loading vulnerability…
CVE-2026-54415 is a missing authorization / broken access control vulnerability in Azuriom CMS affecting versions before 1.2.11 on all platforms. The flaw is present in the server management routes defined in routes/admin.php. An authenticated attacker who has the admin.access permission can send crafted HTTP requests to /admin/servers/create to create AzLink server tokens, then abuse AzLink API endpoints including /api/azlink/password, /api/azlink/email, and /api/azlink/user/{id} to modify non-admin user account attributes. This enables unauthorized password and email changes and results in takeover of targeted non-admin accounts.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos.
This repository is a compact standalone Python proof-of-concept for CVE-2026-54415 affecting Azuriom CMS versions prior to 1.2.11. The repo contains four files: `poc.py` (main exploit), `README.md` (vulnerability write-up, exploit chain, remediation, and usage), `requirements.txt` (`requests` dependency), and `LICENSE` (MIT). The exploit is not part of a larger framework. `poc.py` is the sole code file and clear entry point. It uses `requests.Session()` to perform an authenticated multi-step web exploit: (1) fetch CSRF token and log in to `/user/login` as a low-privileged admin, (2) access `/admin/servers/create` and POST to `/admin/servers` to create a malicious `mc-azlink` server entry, (3) enumerate `/admin/servers` and scrape `/admin/servers/{id}/edit` pages to recover the newly minted 32-character AzLink token from the setup command, and (4) call `/api/azlink/password` with the `Azuriom-Link-Token` header to reset the password of any non-admin user identified by `game_id`. The script then reports successful account takeover. A `--token-only` option allows stopping after token extraction. The exploit’s main capability is authenticated privilege abuse leading to account takeover, not remote code execution. It relies on broken access control in Azuriom’s admin server-management routes, where users with only `admin.access` could create servers before the `admin.servers` permission was introduced in 1.2.11. The PoC specifically abuses the `mc-azlink` server type because the linked verification path does not require a reachable real game server, allowing arbitrary token minting with a dummy address such as `127.0.0.1`. Overall, this is a real, operational exploit PoC with a hardcoded workflow and customizable target credentials/parameters. It is offensive in nature and goes beyond detection by actively changing victim credentials.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
4 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.