MetaGPT hat eine eval injection in metagpt/strategy/tot.py

The impact of this code injection vulnerability is significant. An attacker can leverage the generate_thoughts function to inject malicious code into the MetaGPT system. This could lead to complete system compromise, including data exfiltration, denial of service, or further exploitation of other vulnerabilities. The publicly available exploit dramatically lowers the barrier to entry for attackers, increasing the likelihood of exploitation. Given MetaGPT's potential use in automated tasks and decision-making, the consequences could extend beyond simple data breaches, potentially impacting business operations and strategic decisions.

Organizations utilizing MetaGPT FoundationAgents in production environments, particularly those with limited network segmentation or inadequate input validation practices, are at significant risk. Development teams integrating MetaGPT into their workflows should also be aware of this vulnerability and prioritize patching.

• python / server:

import os
import subprocess

def check_metagpt_version():
    try:
        result = subprocess.check_output(['pip', 'show', 'metagpt'], stderr=subprocess.STDOUT)
        version = result.decode('utf-8').split('Version: ')[1].strip()
        if version <= '0.8.2':
            print(f"MetaGPT version is vulnerable: {version}")
        else:
            print(f"MetaGPT version is patched: {version}")
    except FileNotFoundError:
        print("MetaGPT is not installed.")
    except Exception as e:
        print(f"Error checking MetaGPT version: {e}")

check_metagpt_version()

• python / supply-chain: Monitor Python package dependencies for known vulnerabilities using tools like pip audit or safety. • generic web: Monitor access logs for unusual requests targeting the metagpt/strategy/tot.py endpoint.

1. 2026-04-11Disclosure

   disclosure

2. 2026-04-11PoC

   poc

3. 2026-04-11CISA KEV

   kev

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

The primary mitigation for CVE-2026-6110 is to upgrade MetaGPT FoundationAgents to a version beyond 0.8.2 as soon as a patched version is released by the project maintainers. Until a patch is available, consider implementing strict input validation on any data passed to the generate_thoughts function. While a direct WAF rule is unlikely to be effective without understanding the specific injection patterns, monitoring for unusual code execution within the metagpt/strategy/tot.py file can provide early warning signs. Review and restrict permissions for the user account running the MetaGPT process to limit the potential damage from a successful exploit.