/v1/systemone bypasses the --trust-request-chat-template gate
GHSA-xgh8-qwjh-q853 · published 2026-11-18
- Severity
- Medium
- CVSS score
- 6.5
- Chance of exploitation (EPSS)
- No CVE
- Fixed in
- 0.34.0
A caller that can reach /v1/systemone controls how the conversation is serialized. The system block, which an operator normally pins with --chat-template, can be replaced, and special tokens or turn boundaries that message-level validation would not allow can be injected. Default configuration is affected; no flag is needed. To keep the report accurate, two limits: We are not claiming CPU/memory exhaustion. HfRenderer already bounds a single render with VLLM_CHAT_TEMPLATE_RENDER_TIMEOUT (30 s default, renderers/hf.py:992), and that guard is on this code path too. We have not tested whether the Jinja sandbox can be escaped, and are not claiming it can be. …
Upgrade vllm to 0.34.0 or later. If you start it from an MCP config with npx or uvx, pin that version so an old cached copy is not used.
The issue affects vLLM 0.32.0rc0 through 0.33.0. Version 0.34.0 applies the template trust check before rendering structured-decision requests. A fix for this was introduced in https://github.com/vllm-project/vllm/pull/64503.
| Package | Registry | Affected | Fixed in |
|---|---|---|---|
| vllm | PyPI | >= 0.32.0rc0, < 0.34.0 | 0.34.0 |
- CriticalLiteLLM has SQL Injection in Proxy API key verificationlitellmExploited
- Criticaln8n Vulnerable to Remote Code Execution via Expression Injectionn8n, n8nExploited
- HighLiteLLM has a sandbox escape in custom-code guardraillitellmExploited
- HighLangflow affected by Remote Code Execution via validate_code() exec()langflowExploited
- CriticalJupyter MCP Server: Browser DNS rebinding can hijack the Jupyter backend through management routesjupyter-mcp-server
- Criticalmem0ai Databricks vector store vulnerable to SQL injectionmem0ai
- CriticalvLLM: OpenAI auth bypassvllm
- CriticalvLLM has RCE In Video Processingvllm
- CriticalvLLM Allows Remote Code Execution via PyNcclPipe Communication Servicevllm
- CriticalvLLM Vulnerable to Remote Code Execution via Mooncake Integrationvllm
- CriticalCVE-2025-24357 Malicious model remote code execution fix bypass with PyTorch < 2.6.0vllm
- CriticalvLLM Deserialization of Untrusted Data vulnerabilityvllm
The impact, patch and workaround text comes from the advisory, shortened (published by vLLM). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.