MetaGPT hat ein Injection Problem

Successful exploitation of CVE-2026-5970 allows an attacker to execute arbitrary code on the system running FoundationAgents MetaGPT. This could lead to complete system compromise, including data theft, modification, or destruction. Given the remote nature of the exploit and the availability of a public proof-of-concept, the potential for widespread exploitation is high. The impact is amplified if the MetaGPT instance has access to sensitive data or critical infrastructure, enabling lateral movement and broader network compromise. The lack of a response from the project maintainers further exacerbates the risk.

Organizations and individuals utilizing FoundationAgents MetaGPT versions 0.8.0 through 0.8.1, particularly those deploying it in environments handling sensitive data or critical infrastructure, are at immediate risk. Those relying on MetaGPT for automated code evaluation or testing are especially vulnerable.

• python / server:

import os
import subprocess
# Check for the vulnerable function
with open('/path/to/your/foundationagents/HumanEvalBenchmark/MBPPBenchmark.py', 'r') as f:
    if 'check_solution' in f.read():
        print('Vulnerable function detected!')

• python / supply-chain:

import subprocess
result = subprocess.run(['pip', 'show', 'foundationagents'], capture_output=True, text=True)
if 'Version: 0.8.0' in result.stdout or 'Version: 0.8.1' in result.stdout:
    print('FoundationAgents version is vulnerable!')

1. 2026-04-09Disclosure

   disclosure

2. 2026-04-09PoC

   poc

1. Reserviert2026-04-09
2. Veröffentlicht2026-04-09
3. EPSS aktualisiert2026-04-17

Currently, there is no official patch available for CVE-2026-5970. As a temporary mitigation, consider isolating affected instances of FoundationAgents MetaGPT from external networks to prevent remote exploitation. Input validation and sanitization within the checksolution function could potentially reduce the attack surface, although this requires careful implementation to avoid introducing new vulnerabilities. Monitoring system logs for suspicious activity related to the HumanEvalBenchmark/MBPPBenchmark component is also recommended. After a patch is released, upgrade immediately and confirm the fix by attempting to trigger the checksolution function with malicious input.