EUVD-2026-110038

vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server...

PYSEC-2026-4348

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

PYSEC-2026-4348

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

CVE-2026-115917

The vLLM inference and serving engine for large language models is vulnerable to a Cross-Origin Resource Sharing (CORS) flaw in versions prior to 0.32.0 . The OpenAI-compatible server incorrectly configures Starlette CORSMiddleware with allowed_origins set to a wildcard by default. A remote attac...

CVE-2026-115917 vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

CVE-2026-115917 vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

CVE-2026-115917 vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

CVE-2026-115917

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowedorigins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on...

LiteLLM CVE-2026-47101 Detect

Fingerprints LiteLLM proxy API /health, OpenAI-compatible routes. !/usr/bin/env python3 -- coding: utf-8 -- """Fingerprints LiteLLM proxy API /health, OpenAI-compatible routes.""" from kittysploit import from lib.protocols.http.httpclient import Httpclient class ModuleScanner, Httpclient: info =...

PT-2026-125027

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.32.0 Description The OpenAI-compatible server in this inference engine for large language models configures Starlette CORSMiddleware with allowed origins set to a wildcard by default. This allows a malicious web page t...

CVE-2026-111320

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider baseurl. Attackers can set baseurl via the provider config endpoint and trigger POST /api/v1/models/check to...

CVE-2026-111320 ValueCell through 0.1.20 Unauthenticated SSRF via /api/v1/models/check

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider baseurl. Attackers can set baseurl via the provider config endpoint and trigger POST /api/v1/models/check to...

CVE-2026-111320

ValueCell versions through 0.1.20 are vulnerable to an unauthenticated server-side request forgery (SSRF) vulnerability. The root cause is the use of an unvalidated provider base_url that can be set via the provider configuration endpoint. Remote attackers can exploit this by triggering a POST /a...

CVE-2026-111320

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider baseurl. Attackers can set baseurl via the provider config endpoint and trigger POST /api/v1/models/check to...

CVE-2026-111320 ValueCell through 0.1.20 Unauthenticated SSRF via /api/v1/models/check

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider baseurl. Attackers can set baseurl via the provider config endpoint and trigger POST /api/v1/models/check to...

PT-2026-115278

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider base url. Attackers can set base url via the provider config endpoint and trigger POST /api/v1/models/check to...

CVE-2026-111320: Server-Side Request Forgery (SSRF)

ValueCell through 0.1.20 contains an unauthenticated server-side request forgery vulnerability that allows remote attackers to direct backend requests by setting an unvalidated provider baseurl. Attackers can set baseurl via the provider config endpoint and trigger POST /api/v1/models/check to...

CVE-2026-85668

Xinference affected commit 4a94832, v3.x contains an unauthenticated arbitrary-path file read vulnerability in the POST /v1/models/llm/auto-register endpoint, which accepts a caller-supplied modelpath parameter without authentication or path confinement. The endpoint reads and parses config.json,...

CVE-2026-85668: External Control of File Name or Path

Xinference affected commit 4a94832, v3.x contains an unauthenticated arbitrary-path file read vulnerability in the POST /v1/models/llm/auto-register endpoint, which accepts a caller-supplied modelpath parameter without authentication or path confinement. The endpoint reads and parses config.json,...

LiteLLM Proxy - Model Exposure Detection

LiteLLM proxy was detected with anonymous access to the model listing API. LiteLLM proxies that ship without generalsettings.masterkey or with the key disabled expose the configured upstream model catalog via /v1/models and /model/info — revealing which OpenAI / Anthropic / Azure / Bedrock /...