PYSEC-2026-4220 Headroom vulnerable to Cross-Site WebSocket Hijacking (CSWSH)

Summary The Headroom WebSocket server does not validate the Origin header of incoming client WebSocket requests before forwarding the request to the upstream server, allowing malicious WebSocket clients to perform arbitrary LLM requests without authentication. This can be exploited by a malicious...

EUVD-2026-92842

vLLM: Harmony tool continuations drop cachesalt — restoring a cross-tenant prefix-cache membership oracle...

Use of Cache Containing Sensitive Information

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Use of Cache Containing Sensitive Information via the /v1/responses endpoint, where tokensinput stores the salt on the engine input only when it is...

CVE-2026-105752

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cachesalt value, placing the continuation prefix in the global unsalted cache...

CVE-2026-105752 vLLM: Harmony tool continuations drop `cache_salt` — restoring a cross-tenant prefix-cache membership oracle

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cachesalt value, placing the continuation prefix in the global unsalted cache...

CVE-2026-105752 vLLM: Harmony tool continuations drop `cache_salt` — restoring a cross-tenant prefix-cache membership oracle

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cachesalt value, placing the continuation prefix in the global unsalted cache...

CVE-2026-105752 vLLM: Harmony tool continuations drop `cache_salt` — restoring a cross-tenant prefix-cache membership oracle

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cachesalt value, placing the continuation prefix in the global unsalted cache...

CVE-2026-105752: Exposure of Sensitive Information to an Unauthorized Actor

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cachesalt value, placing the continuation prefix in the global unsalted cache...

PT-2026-106262

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.30.0 Description In vLLM, Harmony tool continuations submitted through the "POST /v1/responses" endpoint rebuild the next-turn engine input without preserving the cache salt value. This causes the continuation prefix t...

GHSA-H46J-26Q3-RGGF Headroom vulnerable to Cross-Site WebSocket Hijacking (CSWSH)

Summary The Headroom WebSocket server does not validate the Origin header of incoming client WebSocket requests before forwarding the request to the upstream server, allowing malicious WebSocket clients to perform arbitrary LLM requests without authentication. This can be exploited by a malicious...

Headroom vulnerable to Cross-Site WebSocket Hijacking (CSWSH)

Summary The Headroom WebSocket server does not validate the Origin header of incoming client WebSocket requests before forwarding the request to the upstream server, allowing malicious WebSocket clients to perform arbitrary LLM requests without authentication. This can be exploited by a malicious...

PT-2026-98928

Name of the Vulnerable Software and Affected Versions BerriAI LiteLLM versions prior to 1.101.0-rc.1 Description A tenant isolation bypass exists in the semantic cache layer due to a metadata key mismatch between the get semantic cache tenant scope and get metadata variable name functions...

CVE-2026-89032: Incorrect Authorization

BerriAI LiteLLM before 1.101.0-rc.1 contains a tenant isolation bypass vulnerability in the semantic cache layer that allows authenticated users to read other tenants' cached responses by exploiting a metadata key mismatch between getsemanticcachetenantscope and getmetadatavariablename. Attackers...

CVE-2026-85666: Server-Side Request Forgery (SSRF)

OGX formerly Llama Stack, affected at commit fbe8e0f contains an unauthenticated server-side request forgery vulnerability in the OpenAI-compatible POST /v1/responses endpoint. MCP tool definitions accept a serverurl parameter along with headers and authorization values that is fetched server-sid...

CVE-2026-73678 MindsDB Minds Platform v26.1.0 Unauthenticated RCE via scratchpad exec()

MindsDB Minds Platform version 26.1.0 and earlier contains an unauthenticated remote code execution vulnerability that allows unauthenticated attackers to execute arbitrary OS commands by submitting crafted prompts to the unprotected POST /api/v1/responses/ endpoint, which reaches the Anton agent...

CVE-2026-73678: Improper Control of Generation of Code

MindsDB Minds Platform version 26.1.0 and earlier contains an unauthenticated remote code execution vulnerability that allows unauthenticated attackers to execute arbitrary OS commands by submitting crafted prompts to the unprotected POST /api/v1/responses/ endpoint, which reaches the Anton agent...

CVE-2026-55255: Authorization Bypass Through User-Controlled Key

Langflow is a tool for building and deploying AI-powered agents and workflows. Prior to 1.9.1, an Insecure Direct Object Reference IDOR vulnerability in /api/v1/responses endpoint allows an authenticated attacker to execute any flow belonging to another user by specifying the victim's flow ID in...