HOL LogoGuard

Explore HOL

  • HOL home
  • AI agent registry
  • AI plugins
  • Open standards
  • HOL members
  • Sponsor HOL
  • Advertise with HOL

Guard product

  • Guard overviewLocal security and control for AI agents and the tools they use.
  • FeaturesRuntime protection, policy routing, review, and evidence.

Explore Guard

  • Product previewWalk through Guard surfaces in read-only demo mode.
  • ComparisonCompare Guard with native controls and AI security vendors.

AI tools

  • All AI toolsEvery supported AI tool and how Guard applies policy to it.
  • Codex
  • Claude Code
  • Cursor
  • Antigravity CLI
  • OpenCode
  • Hermes
  • OpenClaw
  • GitHub Copilot CLI
  • Antigravity
  • Kimi
  • Grok
  • Pi / Oh My Pi
  • Zcode

Extensions

  • All extensionsBrowse command and MCP coverage with owners and stated limits.
  • Command coverageShell command protection across clouds, databases, backups, and packages.
  • MCP server coverageSee how Guard maps risk state across MCP tools and servers.
  • Core safetyThe safety floor listings that ship with Guard.
  • Data and resilienceBackup and storage command protection.
  • Cloud and infrastructureAWS, Azure, GCP, Kubernetes, and more.

Security

  • AI security hubSecurity research, advisories, and agent safety coverage.
  • AI tool securitySecurity profiles for each supported coding agent.
  • Safe labsHands-on attack simulations with safe boundaries.
  • Redacted warningsReal blocked actions with sensitive details removed.
  • AdvisoriesCoordinated disclosure reports for AI tooling.
  • Active CVEsSearch active CVEs affecting AI tooling.

Learn

  • Security guidesPractical guides for securing AI agent workflows.
  • DocsInstall, configure, and operate Guard with confidence.
  • ResearchPublished security research, benchmarks, and methodology.

Community

  • ReleasesVersion history, shipped changes and upgrade notes.
  • ContributorsThe people and contributions behind HOL Guard.
  • AffiliatesShare Guard with your audience and earn from referrals.
  • SponsorFund agreed research, implementation and shared testing through the HOL Partner Program.
  • AdvertiseCompare README logo placements, previews and monthly prices, or promote a plugin or extension.
PricingEnterpriseOpen AppInstall Guard
  1. Guard
  2. Security
  3. CVEs
  4. CVE 2026 115916 vllm out of vocabulary prompt token id passes
HOL Guard

Public security guidance for teams protecting AI harnesses, MCP servers, skills, prompts, and local tool execution.

Install Guard

AI Security

  • Prompt injection
  • MCP security
  • OWASP MCP mapping
  • Supply chain

Resources

  • Trust packet
  • Harness setup
  • Redacted warnings
  • Safe labs

Product

  • Install Guard
  • Pricing
  • Open dashboard
Guard
  • Guard Overview
  • Releases
  • Contributors
  • Install Guard
  • Pricing
Docs
  • Documentation Index
  • Developer Hub
  • API Reference
  • Root OpenAPI
  • Registry OpenAPI
  • Run in Postman
  • Standards
  • Submit ERC-8004 Contract
  • Feature Your Agent
Best Plugins
  • Browse Plugins
  • Plugin Launches
  • Best Claude Plugins
  • Best Codex Plugins
  • Best Grok Plugins
  • Best Kimi Plugins
  • Best DeepSeek Plugins
  • Best Antigravity Plugins
  • Best MCP Servers
  • Best Cursor Plugins
  • Best OpenCode Plugins
Best Agents
  • Best ERC-8004 Agents
  • Best Virtuals Agents
  • Best MCP Servers
  • Best A2A Agents
  • Best x402 Payable
  • All Categories
Community
  • Telegram
  • X
More
  • About HOL
  • Sponsor
  • Advertise
  • Contact
  • Blog
  • GitHub
  • Privacy
  • Terms of Service
Settings

Copyright © 2026 HOL DAO LLC. All rights reserved.

Back to active CVEs
Medium · CVSS 6.5CVE-2026-115916

vLLM: Out-of-vocabulary prompt token id passes input validation and triggers a CUDA device-side assert (engine crash / DoS)CVE-2026-115916

Answer in brief

CVE-2026-115916 records a Medium severity (CVSS 6.5) vulnerability in vLLM: Out-of-vocabulary prompt token id passes input validation and triggers a CUDA device-side assert (engine crash / DoS). The current sources do not mark it as known exploited. The current feed maps vllm-project/vllm (generic). Check affected ranges and fixed versions before updating.

Analysis pending evidence review

HOL Guard separates source facts from reviewed analysis. See the methodology.

Published Oct 26, 2026Updated Oct 26, 2026Source checked Oct 26, 2026First seen by HOL Oct 26, 2026Material review Oct 26, 2026
Upstream Advisory

Key facts

Risk
Medium · CVSS 6.5
Exploitation
Not marked as known exploited
Affected software
1 mapped package or product
Fix availability
Available

Why this deserves its current priority

CVSS is 6.5. The current sources do not mark it as known exploited. Treat this as a source-backed prioritization signal, not a statement about your environment.

Analysis status

Analysis pending evidence review

Factual feed record only; HOL analysis is not approved for indexing. Read the methodology.

Affected scope and exposure questions

The current feed maps vllm-project/vllm (generic). Check affected ranges and fixed versions before updating.

Mapped affected packages and fixed versions
PackageAffected rangeFixed version
vllm-project/vllmgeneric<0.32.00.32.0

Recommended response

  1. 1Check inventory. Check lockfiles and deployed manifests for vllm-project/vllm.
  2. 2Review the reported fix. Update vllm-project/vllm to 0.32.0 if you use the affected versions. Test the change in a non-production environment first.

Evidence timeline and material changes

  1. Published upstream

    Oct 26, 2026

    Evidence: source:cvelist:source_dates:source-dates:record

  2. Source modified

    Oct 26, 2026

    Evidence: source:cvelist:source_dates:source-dates:record

  3. First seen by HOL

    Oct 26, 2026

Sources and claim methodology

  • GitHub security advisorygithub.com
  • GitHub security advisorygithub.com
  • GitHub security advisorygithub.com
  • GitHub security advisorygithub.com
Upstream source description

vLLM is an inference and serving engine for large language models. Prior to 0.32.0, vllm/tokenizers/hf.py computes max_token_id with an off-by-one upper bound, allowing a pre-tokenized prompt ID equal to the vocabulary size to pass validation. For a tensor-parallel size of one and an exact multiple-of-64 vocabulary, VocabParallelEmbedding.forward does not mask that value before F.embedding uses it, causing a CUDA device-side assertion and terminating the engine. This issue is fixed in version 0.32.0.

Quoted source text, attributed separately from HOL analysis.

Related CVEs

  • vLLM: DoS via unbounded chat_template Jinja2 rendering exhausts the renderer thread poolSame vllm-project/vllm package
  • vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabledSame vllm-project/vllm package
  • vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engineSame vllm-project/vllm package
  • vLLM DeepSeek-OCR image processing vulnerable to denial of serviceSame vllm-project/vllm package
  • vLLM: Untrusted EPS inputs can trigger denial of service or Ghostscript RCESame vllm-project/vllm package
  • vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP serverSame vllm-project/vllm package

Record context

Vulnerability class
Secret Exfiltration
EPSS
Not reported
CWE IDs
CWE-129, CWE-193
Source
CVE List V5
Source checked
Oct 26, 2026
References
4 linked sources
Open source record