JunoClaw’s plugin‑shell component contains a substring‑based blocklist that is intended to filter out dangerous shell commands. The blocklist check is performed against the raw command string rather than the first parsed token, enabling an attacker to construct arguments that bypass the filter and cause the host to execute arbitrary shell commands. This flaw effectively provides remote code execution on the system via the plugin‑shell interface.

The vulnerability affects Dragonmonk111:junoclaw versions released prior to the 0.x.y-security-1 release. Users running any earlier release of JunoClaw with the plugin‑shell component are susceptible and should update to 0.x.y-security-1.

The CVSS score of 8.4 indicates severe impact and high exploitability. The vulnerability is not listed in CISA KEV and its exploit probability is currently undefined, but the easily exploitable nature of the blocklist bypass means attackers can readily craft malicious inputs. The likely attack vector is through the plugin‑shell interface, which may be exposed to authenticated users or automated services—an attacker can invoke the plugin with a crafted argument that triggers unauthorized command execution. Applying the 0.x.y-security-1 update mitigates the risk.