CVE-2026-105752
LOW
3,1
Source: [email protected]
Attack Vector: network
Attack Complexity: high
Privileges Required: low
User Interaction: none
Scope: unchanged
Confidentiality: none
Integrity: low
Availability: none
Description
AI Translation Available
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 cache_salt value, placing the continuation prefix in the global unsalted cache namespace even when the caller enabled salting. On deployments with prefix caching enabled, which is the default, an authenticated tenant who can reconstruct a victim's low-entropy post-tool history can submit the same continuation and use the cached_tokens_per_turn count to determine whether the prefix was previously processed, defeating the intended tenant isolation of salted prefix caching. This issue is fixed in version 0.30.0.
200
Exposure of Sensitive Information to an Unauthorized Actor
DraftCommon Consequences
Security Scopes Affected:
Confidentiality
Potential Impacts:
Read Application Data
Applicable Platforms
Technologies:
Not Technology-Specific, Web Based, Mobile
524
Use of Cache Containing Sensitive Information
IncompleteCommon Consequences
Security Scopes Affected:
Confidentiality
Potential Impacts:
Read Application Data
Applicable Platforms
All platforms may be affected
https://github.com/vllm-project/vllm/commit/6a2a2bb02b563b83f946012959fd3927984…
https://github.com/vllm-project/vllm/pull/50195
https://github.com/vllm-project/vllm/pull/51818
https://github.com/vllm-project/vllm/releases/tag/v0.30.0
https://github.com/vllm-project/vllm/security/advisories/GHSA-935w-9g4m-p28p