MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. Each candidate is checked by a language model from its name, description and topics, and only actual MCP servers are listed (not clients, SDKs, lists or skills). Package descriptions, licences and links are read from each package's own registry (npm, PyPI, crates.io, NuGet, RubyGems; Go and Maven packages are only checked to exist), and packages the registry no longer has are removed. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
Tachyon GPU monitorgithub.com/formray/tachyonno update data
Read-only NVIDIA GPU monitor: bottlenecks, VRAM forecast, crash diagnosis, pre-render check, jobs.
mcpb -
mcp-gpu-servergithub.com/mesutoezdil/mcp-gpu-serverupdated this year
MCP server for GPU monitoring: utilization, VRAM, temperature via nvidia-smi / NVML
pypi MIT -
Configurable tile-based system monitor for Windows: per-CPU, processes, GPU, kernel (DPC/ISR per driver, pool tags, objects), an hour of history, plugins and an MCP server for AI agents.
MIT ★ 1 -
Open-source tuning and monitoring for NVIDIA RTX GPUs - overclocking, measured V/F curve undervolting, per-fan curves, RTSS-free FPS metrics, an error-checked stability lab, per-game profiles, and an MCP server so AI agents can tune autonomously. No kernel drivers, no telemetry.
MIT ★ 12 -
Open-source macOS menu bar system monitor for Apple Silicon: live CPU, GPU, memory, power, temperature, and fan. No sudo, no dependencies, with an MCP server for AI agents. A free iStat Menus alternative.
MIT ★ 2 -
coregithub.com/kubeara/coreactive
Self-hosted PaaS for private infrastructure. Deploy 200+ services on any server in one click. GPU monitoring, AI model management, VRAM checker, and MCP built in. Coolify alternative. MIT licensed. Free forever.
MIT ★ 43 -
Shipyard 🚢 — Cross-platform Docker container management platform. Python FastAPI backend + Flutter frontend for Android, iOS, macOS, Web & OpenHarmony. Manage multiple Docker hosts, monitor CPU/GPU resources in real time, control containers & Compose projects, built-in MCP server.
MIT ★ 1 -
AppControl MCP is a local server connecting AI agents and IDEs like Claude, Cursor, and Windsurf to system history provided by the AppControl desktop app. Ask which apps used CPU, GPU, RAM, or disk, and what caused slowdowns, heat spikes, or webcam, mic, and location access. Built for deep historical monitoring and useful real-world PC insights.
go MIT ★ 45 -
MCP server for interacting with Kubernetes clusters via kubectl with NVIDIA GPU monitoring support
npm MIT -
remote-sandbox-mcpgithub.com/DebinXiangQuantum/remote-sandbox-mcpupdated this year
MCP server for SSH-based remote sandboxes: multi-server management, persistent sessions, background tmux tasks, and GPU/CPU/memory resource monitoring
pypi MIT ★ 1 -
Your Mac at a glance. A native menu bar system monitor with CPU, GPU, Neural Engine, power, thermals, fans, per-app volume, dev servers, coding agents, AI usage limits, a CLI and an MCP server.
MIT ★ 2 -
mcp-os-doctorgithub.com/brenordv/mcp-os-doctorupdated this year
A read-only MCP server that lets AI assistants diagnose and monitor Windows system health (hardware sensors), processes, services, event logs, GPU, and DirectX. All without granting write access or command execution.
GPL-3.0 -
eth-euler-hpc-mcpgithub.com/gbrlpzz/eth-euler-hpc-mcpupdated this year
Local MCP server, CLI, and agent skill for safe Slurm job management on ETH Zürich's Euler HPC cluster (free public share). Submit, monitor, cancel, and retrieve jobs; check quotas; transfer data. 48 cores / 128 GiB / no GPUs enforced.
MIT ★ 2 -
Self-hosted homelab dashboard in one container — GPU, local-AI VRAM, Docker, systemd, energy costs. Read-only MCP server so AI agents can explore it.
MIT ★ 237 -
One terminal for the GPUs of all your ssh hosts, with an MCP server so agents can find a free GPU. The remote side needs only nvidia-smi.
MIT ★ 1 -
MCP-Serversgithub.com/saviornt/MCP-Serversupdated this year
Cross-platform MCP telemetry server providing structured system + GPU + network observability tools for AI agents (LM Studio / FastMCP / GHCR).
Apache-2.0 -
cluster-job-monitorgithub.com/dafidofff/cluster-job-monitorupdated this year
Read-only multi-cluster SLURM dashboard + agent capacity overview, over CLI and MCP.
pypi MIT -
117 MCP tools for GPU infrastructure + Agent FinOps — deploy LLMs, manage VMs (Proxmox/XO/vSphere), track cost per agent, enforce budgets and model policies. Works with Claude, Cursor, n8n, LangChain
pypi MIT ★ 20 -
Give Claude Code, Codex and Gemini every machine you have: a live GPU, VRAM, RAM and disk inventory across your servers, SSH access between them, and an MCP server.
pypi MIT -
MCP server (zero-dep Python) to control GPUs via LACT on Linux — stats, power, fan, clocks, voltage/VF curve for AI agents
-
14 checks, one command, graded findings. A health doctor for a home AI stack: Ollama, NVIDIA GPUs, tailnet.
pypi MIT -
Cross-platform hardware inspection, monitoring, stress testing and tuning, with a CLI, TUI and MCP server.
Apache-2.0 -
Real-time and historical metrics for locally served LLMs (Ollama and vLLM) — dashboard, HTTP API and MCP server. One process, one SQLite file, never in the request path.
Apache-2.0 ★ 1 -
CPU and GPU stress test for Linux (GTK 4 / libadwaita) – live sensors and 30-minute graphs, power supply readings, safety limits, Vifte-aware, status icon and Claude (MCP) integration
GPL-3.0 -
Your training runs, live, on your own machines. A self-hosted ML run tracker: phone alerts when a run dies, every GPU in view, MCP for coding agents.
Apache-2.0 ★ 1 -
A Task Manager for the GPU on Windows: per-process VRAM (resident vs committed) + engine load. Treemap GUI, live TUI, JSON/NDJSON, MCP server, co-tenancy stamps for agents. C++20, zero deps, single exe.
MIT -
NVIDIA Triton Inference Server control plane — model repository manipulation, VRAM monitoring, and inference over gRPC via FastMCP 3.2
MIT ★ 1 -
Tiny Windows service that serves a live hardware dashboard (CPU, every NVIDIA GPU, RAM, temperatures, fans) to any browser on your network, with an MCP server for AI agents and mod support
MIT ★ 2