All reports
High

openstack-mistral: mistral: OS command injection via unvalidated proxy_command in std.ssh_proxied action

Affected versions

Affected versions
Unknown
Fixed version
Not reported
Affected versions
Unknown
Fixed version
Not reported
Affected versions
Unknown
Fixed version
Not reported

Exploitation

CVE-2026-93858

Probability of exploitation in the next 30 days: 0.43%

FIRST EPSS · 2026-11-19

Description

In OpenStack Mistral through 23.0.0, the std.ssh_proxied action passes a caller-supplied proxy_command value directly to paramiko.ProxyCommand() before any SSH connection to a gateway or target host is attempted. An authenticated project member can use the standard action-execution API to submit an arbitrary local command as proxy_command; paramiko starts that command as a subprocess on the executor host under the executor's own service account, independent of whether the SSH connection itself ever succeeds. Only Mistral deployments that permit the std.ssh_proxied action, the default configuration, are affected.

A flaw was found in OpenStack Mistral. The std.ssh_proxied built-in action accepts a user-supplied proxy_command parameter that is passed without validation to a system subprocess call on the Mistral executor host. An authenticated attacker can craft a malicious workflow that supplies an arbitrary command string as the proxy_command parameter, causing the Mistral executor to run that command as its operating system process user. Successful exploitation allows full remote code execution on the executor host with the privileges of the Mistral executor service account.