The vulnerability resides in FoundationAgents MetaGPT’s `check_cmd_exists` function, where manipulation of the `mermaid.path` argument allows an attacker to inject arbitrary shell commands. This defect can lead to execution of any OS command within the context in which the application runs, effectively compromising confidentiality, integrity, and availability of the affected system.

The issue affects FoundationAgents’ MetaGPT product up to version 0.8.2. Users running these releases, irrespective of operating system, are at risk if the `check_cmd_exists` function processes untrusted input for the `mermaid.path` setting.

The CVSS score of 2.3 indicates a low severity rating, and the EPSS score is not available, suggesting limited exploitation data. The vulnerability can be triggered remotely, but the attac­k requires a high degree of complexity and is known to be difficult to exploit. The CVE is not listed in CISA’s Known Exploited Vulnerabilities catalog. Nonetheless, the publicly disclosed exploit indicates potential risk, especially if the application is exposed to untrusted users or networks. Monitoring for official patches remains prudent.