These pages are published after PatchSiren validates generated defensive summaries against stored public CVE and source evidence.
A critical vulnerability was found in the new API, a large language model (LLM) gateway and artificial intelligence (AI) asset management system. Prior to version 1.0.0-rc.18, user-controlled input could lead to overflow conversions, allowing a low-privileged account to potentially drain upstream funds by turning a negative charge into account credit.
CVE-2026-64868 is a vulnerability in an API gateway and AI asset management system, allowing unauthenticated attackers to cause memory pressure, container restarts, or disk exhaustion. The issue was fixed in version 1.0.0-rc.11. This vulnerability affects the system's ability to handle request bodies before signature validation, potentially leading to resource pressure. Defenders should assess exposure an [truncated]
CVE-2026-64866 is a vulnerability in the AdminResetPasskey function of an LLM gateway and AI asset management system, allowing lower-privileged administrators to remove passkeys from higher-privileged accounts, including root accounts. The issue was fixed in version 1.0.0-rc.7. This vulnerability could lead to unauthorized privilege escalation and increased risk of lateral movement within multi-administra [truncated]
CVE-2026-64865 is a race condition vulnerability in the new API's large language model (LLM) gateway and artificial intelligence (AI) asset management system. An authenticated user can exploit this issue to artificially inflate their cached quota by repeatedly sending PUT requests to /api/user/self, which updates language or sidebar modules. This occurs because the controller/user.go file calls User.Updat [truncated]
CVE-2026-64859 is a critical vulnerability in an LLM gateway and AI asset management system, allowing authenticated administrators to obtain the root user's bearer token and access root-only system configuration APIs. The issue is fixed in version 1.0.0-rc.7. Affected product deployments should be verified, and administrators should review official advisories to validate scope and severity. Compensating c [truncated]
A vulnerability was discovered in the New API large language mode (LLM) gateway and artificial intelligence (AI) asset management system, prior to version 0.12.0-alpha.1. The email and WeChat account binding endpoints, GET /api/oauth/email/bind and GET /api/oauth/wechat/bind, used GET requests for state-changing account operations. This allowed an attacker to trigger a logged-in user's browser to bind an [truncated]
CVE-2026-33655 is a high-severity SSRF vulnerability in an AI asset management system. Prior to version 0.12.0-alpha.1, the default SSRF protection configuration did not apply IP filtering to hostnames. With ApplyIPFilterForDomain disabled by default, URL validation checked domain allow/block rules but did not resolve a hostname and validate the resolved IP address. This allowed authenticated users to con [truncated]
A low-severity authorization bypass vulnerability affects QuantumNous new-api versions up to 0.12.1. The flaw resides in the RelayMidjourneyImage/GetByOnlyMJId function within router/relay-router.go, specifically at the Midjourney Image Relay Endpoint. Successful exploitation allows remote attackers to bypass authorization controls, though the attack requires high complexity and is considered difficult to [truncated]
A SQL injection vulnerability exists in QuantumNous new-api versions up to 0.12.1. The vulnerability is located in the SearchUserTopUps and SearchAllTopUps functions within model/topup.go. An authenticated attacker with low privileges can exploit this remotely to manipulate database queries. The CVSS 4.0 score of 2.1 reflects limited impact scope, though the public availability of exploit information incr [truncated]
New API is an LLM gateway and AI asset management system. Prior to version 0.12.10, the Stripe webhook handler fails to properly validate webhook event authenticity, allowing unauthenticated attackers to forge webhook events and credit arbitrary quota to their accounts without payment. The vulnerability stems from insufficient verification of webhook signatures (CWE-345: Insufficient Verification of Data [truncated]