LangChain CommunityにおけるSSRF脆弱性 (SSRF vulnerability) がRequestsToolkitコンポーネントに存在します

The SSRF vulnerability in langchain-community allows an attacker to craft malicious requests that the application will execute on the server-side. This can lead to several serious consequences. Attackers could perform internal port scans to map the network infrastructure, potentially identifying vulnerable services. More critically, they can access local services that are not exposed to the public internet, such as internal APIs or databases. In cloud environments like AWS or Azure, the attacker could retrieve instance metadata, exposing sensitive information like API keys and credentials. This could facilitate further lateral movement and compromise of the entire system. The ability to interact with internal resources makes this a high-impact vulnerability.

Organizations deploying langchain-community in production environments, particularly those utilizing cloud-based infrastructure, are at significant risk. Applications that rely on langchain-community for interacting with external services or internal APIs are also vulnerable. Shared hosting environments where multiple users share the same server instance are particularly susceptible.

• python / server:

import langchain
import requests

# Check langchain version
print(langchain.__version__)

# Attempt a request to a local address (e.g., 127.0.0.1:8080) to test SSRF
# This is a simplified example and may require adjustments based on the application's context
try:
    response = requests.get('http://127.0.0.1:8080')
    print(f'Request successful: {response.status_code}')
except requests.exceptions.RequestException as e:
    print(f'Request failed: {e}')

1. 2025-06-23Disclosure

   disclosure

1. 予約済み2025-03-26
2. 公開日2025-06-23
3. 更新日2025-07-17
4. EPSS 更新日2026-03-23

The primary mitigation for CVE-2025-2828 is to immediately upgrade the langchain-community package to version 0.0.28 or later. This version includes a fix that restricts requests to remote internet addresses. If upgrading is not immediately feasible, consider implementing a Web Application Firewall (WAF) or proxy to filter outbound requests and block those targeting internal IP addresses or sensitive endpoints. Additionally, review and restrict the permissions granted to the langchain application to minimize the potential impact of a successful SSRF attack. Carefully examine the RequestsToolkit configuration to ensure it adheres to the principle of least privilege.