About this integration
Read-only ops/health MCP server for a local vLLM instance: health tiers, GPU/VRAM, service status.
- Transport
- stdio
- Authentication
- none
- Initial setup
- none
- Runtime
- unattended
- Evidence
- documented
- Version
- 0.1.0
- Package
- vllm-ops-mcp
- Last compatibility test
- Not independently tested
Connect your agent
vllm-ops-mcpPublisher README and Python manifest reviewed; package, vLLM, GPU, WSL, and systemd were not executed. Qualification covers the documented local default stdio setup with no environment overrides. Deep health and test-completion tools perform real inference and are rate-limited; no lifecycle-control tool is present. Other vLLM deployments, authentication layers, and multi-instance setups need separate review.
Capabilities: Check local vLLM health, List served models, Run bounded test completions, Inspect GPU and service status, Inspect vLLM launch configuration
Connected profiles
Additional details
io.github.jaimenbell/vllm-ops-mcp
Source ↗ · Checked 2026-09-170.1.0
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17pypi
Source ↗ · Checked 2026-09-17vllm-ops-mcp
Source ↗ · Checked 2026-09-190.1.0
Source ↗ · Checked 2026-09-19The publisher documents registration as a local stdio process with its own environment and no overrides for the stated default setup.
Source ↗ · Checked 2026-09-19The reviewed localhost default documents no environment overrides or credential setting; this no-auth classification does not extend to separately secured vLLM deployments.
Source ↗ · Checked 2026-09-19The MCP client configuration launches the local server with stdio transport.
Source ↗ · Checked 2026-09-19Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-19