Security: BerriAI/litellm
Security Advisories
View known security vulnerabilities and report new vulnerabilities privately to maintainers.
-
Local file read via request-supplied OIDC file referencesGHSA-4g5m-c9r5-49xf published
Jun 30, 2026 by jaydnsLow -
Arbitrary file write via path traversal in Skills archive extractionGHSA-5jmr-gcrj-2c9q published
Jun 30, 2026 by jaydnsModerate -
Authentication Bypass via Host Header InjectionGHSA-4xpc-pv4p-pm3w published
May 28, 2026 by jaydnsCritical -
Custom Code Guardrails production endpoints bypass code safety checksGHSA-72m8-9m7m-h278 published
Jun 30, 2026 by jaydnsLow -
MCP Authentication Bypass via OAuth2 Passthrough FallbackGHSA-7488-6r32-c95q published
Jun 30, 2026 by jaydnsHigh -
Authenticated command execution via MCP stdio test endpointsGHSA-v4p8-mg3p-g94g published
Apr 21, 2026 by jaydnsHigh -
Sandbox escape in custom-code guardrailGHSA-wxxx-gvqv-xp7p published
May 7, 2026 by jaydnsModerate -
SQL injection in Proxy API key verificationGHSA-r75f-5x8p-qvmc published
Apr 20, 2026 by jaydnsCritical -
Server-Side Template Injection in /prompts/test endpointGHSA-xqmj-j6mv-4862 published
Apr 20, 2026 by jaydnsHigh -
Password hash exposure and pass-the-hash authentication bypassGHSA-69x8-hrgq-fjj8 published
Apr 4, 2026 by jaydnsHigh