Foundation Agents MetaGPT Deserialization of Untrusted Data Remote Code Execution Vulnerability

A remote code execution vulnerability has been identified in Foundation Agents MetaGPT. This issue arises from the deserialize_message function, where user-supplied data is not properly validated, allowing for the deserialization of untrusted data. As a result, remote attackers can execute arbitrary code on affected installations, using the service account's privileges. Notably, no authentication is required to exploit this vulnerability.

Impact

Exploitation of this vulnerability allows for arbitrary code execution on the affected system, with the executed code running in the context of the service account.

Remediation

Given the nature of the vulnerability, the primary mitigation strategy is to limit interaction with the product.