These pages are published after PatchSiren validates generated defensive summaries against stored public CVE and source evidence.
CVE-2026-54040 is a medium-severity vulnerability in LibreChat, a ChatGPT clone, that allows an attacker to regenerate 2FA backup codes without verification, potentially bypassing 2FA login or disabling it entirely. This issue was fixed in version 0.8.4-rc1. The vulnerability has a CVSS score of 5.9 and is considered a significant risk. Users of LibreChat should ensure they are running version 0.8.4-rc1 o [truncated]
CVE-2026-54030 is a high-severity vulnerability in LibreChat, an enhanced ChatGPT clone. The issue lies in the MCP OAuth implementation, which fails to validate the resource parameter from OAuth Protected Resource metadata. This oversight enables a malicious MCP server to steal access tokens intended for a legitimate server. The vulnerability was publicly disclosed on June 25, 2026, and was fixed in Libre [truncated]
CVE-2026-54029 is a medium-severity vulnerability in LibreChat, an enhanced ChatGPT clone. The vulnerability allows authenticated users to delete other users' messages. This is possible because the DELETE /api/messages/:conversationId/:messageId endpoint does not properly validate the requesting user's permissions. An attacker can exploit this by providing a valid conversationId and the victim's messageId [truncated]
A vulnerability in LibreChat's markdown artifact preview pipeline allows attackers to inject arbitrary event handlers. The marked library v15.0.12 does not HTML-escape double-quote characters in image alt text when a custom renderer falls through to the default renderer. LibreChat's generateMarkdownHtml function installs a custom image renderer that returns false for URLs passing the isSafeUrl allowlist c [truncated]
CVE-2026-54024 is a vulnerability in LibreChat, an enhanced ChatGPT clone, that allows authenticated users to upload arbitrarily large files. This is possible because the POST /api/convos/import endpoint uses a separate multer instance that was not updated with the same limits configuration as the fix for CVE-2024-11171. The application-level size check is also disabled by default. An attacker could explo [truncated]
CVE-2026-34371 is an arbitrary file write vulnerability in LibreChat, a ChatGPT clone, prior to version 0.8.4. The execute_code sandbox feature trusts user-inputted filenames, allowing malicious users to write files to arbitrary locations on the server using traversal sequences. This vulnerability can be exploited by providing a malicious filename containing traversal sequences, which is then concatenated [truncated]
