FoundationAgents MetaGPT versions 0.8.0 and 0.8.1 contain a server-side request forgery vulnerability in the decode_image function of metagpt/utils/common.py. This vulnerability arises from improper validation or sanitization of the img_url_or_b64 parameter, enabling remote attackers to coerce the server into making arbitrary HTTP requests. The vulnerability is remotely exploitable without user interaction and has a CVSS 4.0 base score of 5.3, indicating medium severity. The project has been notified but has not issued a fix or mitigation guidance. Exploit code has been publicly released.

Successful exploitation of this vulnerability allows an attacker to make the vulnerable server perform arbitrary HTTP requests, potentially accessing internal resources or services not otherwise exposed. This can lead to information disclosure or facilitate further attacks within the internal network. The medium CVSS score reflects limited impact due to required privileges and low complexity of attack, but the availability of public exploits increases risk.

No official fix or patch is currently available from the vendor. Users should monitor the FoundationAgents MetaGPT project for updates and apply patches once released. Until then, consider restricting network access from the MetaGPT server to trusted endpoints only, and avoid processing untrusted input for the decode_image function. Patch status is not yet confirmed — check the vendor advisory for current remediation guidance.