Machine-written summary
Summary written by Claude Haiku from the source advisory. Not reviewed by a person.
vLLM versions before 0.30.0 are affected. On flash late-interaction `/score` and `/rerank` deployments, which are enabled by default for supported models, the server caches query embeddings under a key derived from the caller-supplied X-Request-Id header. A remote client who reuses another request's header value can have the victim's documents scored against the attacker's query, or cause the victim's request to fail with a cache-miss error. Upgrade to vLLM 0.30.0 or later.
Not analysed yet
We haven’t published our analysis of this Advisory yet. Once a person has reviewed it, this page will say what’s vulnerable, what has to be true for your app to be affected, and how to check.
Severity and exploitation
| Score | Rating | CVSS | Vector | Source |
|---|---|---|---|---|
| 4.2 | Medium | 3.1 | CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:L | CNA (GitHub_M) |
- CISA KEV Not listed.
- EPSS 0.0020, the 10th percentile, on 8 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.
Timeline
- 31 Oct 2026, 14:08 UTC First seen by agentprime
- 31 Oct 2026, 14:08 UTC Confirmed by OSV
Sources
- GitHub Advisory Database: GHSA-2phq-3phc-84px
- OSV: GHSA-2phq-3phc-84px
- github.com/vllm-project/vllm/security/advisories/GHSA-2phq-3phc-84px
- github.com/vllm-project/vllm/pull/51445
- github.com/vllm-project/vllm/commit/ee17d0d869203ef9a35ad73358a4987bba14b1fc
- github.com/vllm-project/vllm
- github.com/vllm-project/vllm/releases/tag/v0.30.0