Advisory record · GHSA-2pc9-4j83-qjmr

vLLM affected by RCE via auto_map dynamic module loading during model initialization

CVSS 8.8 HighEPSS 0.0098 · 62nd percentile

Published by the GitHub Advisory Database at 21 Jan 2026, 16:12 UTC. agentprime has not analysed this yet, so whether it affects any app is not known.

PackageEcosystemAffectedFixed in
vllmPyPI>=0.10.1, <0.14.00.14.0
Identifiers
GHSA-2pc9-4j83-qjmr · CVE-2026-22807 · PYSEC-2026-2010
Weakness (CWE)
CWE-94
Published
21 Jan 2026
Updated
10 Sept 2026

Machine-written summary

Summary written by Claude Haiku from the source advisory. Not reviewed by a person.

vLLM versions 0.10.1 up to but not including 0.14.0 are affected. An attacker who can influence the model repository or path, whether a local directory or a remote Hugging Face repo, can run arbitrary Python code on the vLLM host during model loading, even when trust_remote_code is false. No API access is needed. Upgrade to vLLM 0.14.0 or later.

Written 5 Nov 2026.

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

ScoreRatingCVSSVectorSource
8.8 High 3.1 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H CNA (GitHub_M)
  • CISA KEV Not listed.
  • EPSS 0.0098, the 62nd percentile, on 14 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.

Timeline

  1. 31 Oct 2026, 14:36 UTC First seen by agentprime
  2. 31 Oct 2026, 14:36 UTC Confirmed by OSV

Sources

Facts on this page come from the GitHub Advisory Database (CC-BY 4.0), OSV and the other sources listed, as published. We don’t reproduce full advisory descriptions, which can contain exploit code.