Ridvay Code’s auto‑approval module contains an OS command injection flaw that completely bypasses its whitelist protection. The module uses fragile regular expressions to parse command structures, but fails to account for standard shell command substitution such as $(...) and backticks. An attacker can craft a command like git log --grep="$(malicious_command)"; the system misidentifies it as a legitimate git operation and automatically approves it, causing the underlying shell to execute the injected code. This results in remote code execution without any user interaction.

The vulnerability appears in the Ridvay Code Auto‑Approval Module. No version information is provided, implying the flaw may exist across all released updates until a patch is released.

The CVSS score of 9.8 indicates extremely high severity. EPSS is below 1%, so the statistical likelihood of exploitation is currently low, and the issue is not listed in CISA’s KEV catalog. The attack vector is inferred to be remote, relying on an attacker’s ability to submit a command to the auto‑approval interface, possibly through a web form or API. Once the injection succeeds, the attacker gains full control over the host system, compromising confidentiality, integrity, and availability. Because no user interaction is required, exploitation could occur in unattended environments, increasing risk to unattended servers or CI/CD pipelines.