The flaw resides in the handling of the sentinel_token setting in Coolify versions prior to v4.0.0-beta.466. When the token is passed to shell commands without sanitization, an attacker who can modify that setting can inject shell syntax. The resulting command execution occurs on the host when Sentinel is restarted, which is a form of remote code execution (CWE-78).

All deployments of CoollabsIO Coolify before version 4.0.0-beta.466 are affected. The vulnerability arises from unvalidated input in the sentinel_token configuration that is used by the application’s internal shell invocations. Users with permission to edit Sentinel settings or to restart Sentinel on the server are therefore at risk.

The CVSS score of 8.8 indicates high severity, while the EPSS score of < 1% suggests a very low but non‑zero likelihood of exploitation. The vulnerability requires authenticated access, meaning an authorized or hijacked user can gain host‑level control. Although not yet listed in the CISA KEV catalog, the attack is feasible by granting the user permission to modify Sentinel settings and then triggering a Sentinel restart, which runs the injected shell syntax under the service user’s privileges.