OAuth Vulnerability in LibreChat Affects Multiple AI Provider Integrations
CVE-2026-54031
6.3MEDIUM
What is CVE-2026-54031?
An OAuth vulnerability in LibreChat allows attackers to exploit redirect-following fetch behavior without proper validation of redirect targets, potentially enabling redirection of requests to internal or private services. The issue arises from a flaw in how initial destination validation is conducted for MCP OAuth requests. This flaw can be notably detrimental if an attacker controls the OAuth endpoint, as it can bypass expected security measures, leading to unauthorized access to internal services. This vulnerability was resolved in version 0.8.8, enhancing the application's security posture.
Affected Version(s)
LibreChat < 0.8.8
