Remote Code Execution in infiniflow/ragflow

The impact of CVE-2024-10131 is severe. An attacker can leverage the vulnerability to execute arbitrary code within the context of the ragflow application. This could lead to complete system compromise, including data exfiltration, malware installation, and denial of service. The ability to dynamically instantiate classes based on user input bypasses intended security controls, creating a direct pathway for malicious code execution. Successful exploitation could grant an attacker persistent access and control over the affected system, potentially leading to broader network compromise if the system has access to sensitive resources or internal networks.

Organizations deploying ragflow in production environments, particularly those with user-facing interfaces that accept LLM factory configurations, are at significant risk. Development teams using ragflow for rapid prototyping or experimentation should also prioritize patching to prevent accidental exposure of vulnerable code. Shared hosting environments where multiple users can potentially influence application configuration are especially vulnerable.

• python / application:

import os
import subprocess

def check_ragflow_version():
    try:
        result = subprocess.run(['pip', 'show', 'ragflow'], capture_output=True, text=True, check=True)
        version = result.stdout.split('Version: ')[1].split('\n')[0]
        if version <= '0.11.0':
            print(f"Vulnerable version detected: {version}")
        else:
            print(f"Safe version detected: {version}")
    except subprocess.CalledProcessError:
        print("ragflow not installed.")

check_ragflow_version()

• linux / server:

ps aux | grep -i ragflow

• generic web: Inspect HTTP requests to the addllm endpoint for suspicious values in the llmfactory parameter. Look for unusual characters or patterns that could indicate an attempt to inject malicious code.

1. 2024-10-19Disclosure

   disclosure

1. 予約済み2024-10-18
2. 公開日2024-10-19
3. 更新日2025-10-15
4. EPSS 更新日2026-04-02

The primary mitigation for CVE-2024-10131 is to upgrade to a patched version of ragflow as soon as it becomes available. Until a patch is applied, implement strict input validation on all user-supplied data, particularly the req['llmfactory'] and req['llmname'] parameters within the add_llm function. Consider implementing a Web Application Firewall (WAF) with rules to detect and block suspicious requests containing potentially malicious payloads. Regularly review and audit the ragflow configuration to ensure adherence to security best practices. Monitor application logs for any unusual activity or error messages related to class instantiation.