vllm-ops-mcp
A read-only operations and health monitoring MCP server designed for local vLLM instances running on WSL2 or systemd. It provides tools to check liveness and real-completion health tiers, monitor GPU and VRAM usage, inspect systemd service status, and introspect live serve flags. It is intended for developers managing their own local AI inference servers without requiring container lifecycle management.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Check vLLM server liveness and real-completion health
- Monitor GPU utilization and VRAM usage
- Inspect systemd service status for vLLM
- Perform manual sanity check completions
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
vllm-ops-mcp 0.1.0
{
"version": "0.1.0",
"transport": "stdio",
"identifier": "vllm-ops-mcp",
"registryType": "pypi"
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"vllm-ops-mcp": {
"command": "uvx",
"args": [
"vllm-ops-mcp==0.1.0"
]
}
}
}Cline
{
"mcpServers": {
"vllm-ops-mcp": {
"command": "uvx",
"args": [
"vllm-ops-mcp==0.1.0"
]
}
}
}VS Code
{
"servers": {
"vllm-ops-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"vllm-ops-mcp==0.1.0"
]
}
}
}Claude Desktop
{
"mcpServers": {
"vllm-ops-mcp": {
"command": "uvx",
"args": [
"vllm-ops-mcp==0.1.0"
]
}
}
}Let your agent find its tools
Find servers like vllm-ops-mcp from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.