MetaGPT hat eine eval-Injektion über einen Cross-Site Request Forgery Angriff

The CSRF vulnerability in MetaGPT FoundationAgents allows an attacker to craft malicious requests that appear to originate from a legitimate user. This can be leveraged to perform actions such as modifying configurations, executing commands, or accessing sensitive data within the MetaGPT environment. The impact is amplified if the user has elevated privileges or if the system interacts with other critical services. While the specific attack surface depends on the application's functionality, the potential for unauthorized actions is significant. The publicly disclosed nature of this vulnerability increases the risk of exploitation.

Organizations and individuals utilizing MetaGPT FoundationAgents versions 0.8.2 and earlier, particularly those integrating the Mineflayer HTTP API with other systems, are at significant risk. Shared hosting environments where multiple users share the same MetaGPT instance are especially vulnerable, as an attacker could potentially exploit the vulnerability on behalf of other users.

• python / server:

import requests
from bs4 import BeautifulSoup

# Example: Check for suspicious requests to the API
url = "http://your-metagpt-instance/api/mineflayer"
response = requests.get(url)

if response.status_code == 200:
    soup = BeautifulSoup(response.content, 'html.parser')
    # Look for unexpected parameters or actions
    if soup.find('param', {'name': 'malicious_action'}) is not None:
        print("Potential CSRF attack detected!")

• generic web:

curl -I http://your-metagpt-instance/api/mineflayer | grep -i 'referer'

1. 2026-04-11Disclosure

   disclosure

1. Reserviert2026-04-11
2. Veröffentlicht2026-04-11
3. Geändert2026-04-14
4. EPSS aktualisiert2026-04-19

Due to the lack of a released patch, immediate mitigation strategies are crucial. Implement strict input validation and output encoding to prevent malicious data from being processed. Consider implementing CSRF protection mechanisms, such as synchronizer tokens or double-submit cookies, to mitigate the risk of unauthorized requests. Restrict access to sensitive endpoints and implement robust authentication and authorization controls. Monitor network traffic for suspicious activity and consider using a Web Application Firewall (WAF) to filter out malicious requests. Regularly review and update the MetaGPT FoundationAgents environment to address potential vulnerabilities.