vllm

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a reference guide for using the vLLM inference engine, containing only documentation links and operational rules.
  • [SAFE]: No external dependencies, package installations, or remote code execution patterns were identified.
  • [SAFE]: The skill mentions using a 'dummy key' for local server connections, which is a standard configuration for non-authenticated local inference endpoints and does not represent a credential leak.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 11:52 PM
Security Audit — agent-trust-hub — vllm