The SitemapLoader in langchain-community is susceptible to Server-Side Request Forgery, allowing unauthenticated attackers to probe internal network resources.

The SitemapLoader fails to enforce domain restrictions on nested sitemap index entries, which allows for Server-Side Request Forgery (CWE-918) that can be triggered by an unauthenticated attacker.

An attacker can use this vulnerability to perform requests against internal services or cloud metadata endpoints that are otherwise inaccessible from the public internet. This exposure can lead to the discovery of internal network topology and potential unauthorized access to sensitive internal data, justifying the high CVSS score of 8.6.

Immediate Action: Review the official LangChain security advisories and update to the latest version of langchain-community where the SitemapLoader domain validation logic has been hardened.

Proactive Monitoring: Monitor outbound network traffic from the application host for requests directed toward internal IP addresses or sensitive infrastructure endpoints.

Compensating Controls: Use network-level egress filtering to restrict the application from reaching internal network segments or unauthorized external domains.

Public Exploit Available: No

Due to the risk of internal network reconnaissance and potential data leakage, organizations using the SitemapLoader component should prioritize updating their libraries. Implementing egress filtering is a highly recommended secondary measure to minimize the blast radius of potential SSRF vulnerabilities.