Cloud & Infrastructure MCP servers connect AI models directly to cloud providers, container orchestrators, and terminal environments. They enable assistants to run shell commands, interact with AWS S3 storage buckets, provision RunPod GPU instances, manage domains, and inspect Kubernetes clusters safely from local or remote developer tooling.
Cloud & Infrastructure MCP servers let AI assistants inspect, provision, and maintain computing resources, cloud storage, container orchestrators, and system shells directly from LLM interfaces. By standardizing communication through the Model Context Protocol, these servers bridge conversational models to developer environments, reducing the need to manually toggle between terminal windows, web consoles, and deployment dashboards.
Notable options include mcp-k8s-go, a Golang-based server for managing Kubernetes clusters; Shell MCP Server, which handles shell command execution with configurable directory and timeout controls; RunPod MCP Server, designed to query and control GPU cloud resources through the RunPod REST API; and s3-tools MCP Server, which supplies dedicated operations for interacting with AWS S3 buckets.
| MCP server | What it connects to | Type | Repository |
|---|---|---|---|
| ShellAgent | ShellAgent is an MCP server that provides terminal command execution capabilities to AI clients and coding assistants. It connects Model | Open Source | https://github.com/myshell-joe/Shellagent-MCP |
| mcp-k8s-go | mcp-k8s-go is an MCP server that provides Large Language Model clients direct access to Kubernetes clusters using Go-based tools and | Open Source | https://github.com/strowk/mcp-k8s-go |
| Shell MCP Server | Shell MCP Server is an MCP server that enables AI clients to execute command-line operations securely across local file systems. | Unknown | https://github.com/blazickjp/shell-mcp-server |
| Sherlock Domains | Sherlock Domains is an MCP server that provides tools to buy and manage web domains alongside a built-in note storage | Unknown | https://github.com/Fewsats/sherlock-mcp |
| Shipyard | Shipyard is an MCP server that integrates the Shipyard CLI with Model Context Protocol clients to inspect and manage ephemeral | Unknown | https://github.com/shipyard/shipyard-cli |
| Shell Executor | Shell Executor is an MCP server that provides controlled command-line execution capabilities to Model Context Protocol clients like VS Code | Open Source | https://github.com/yourusername/shell-executor-mcp |
| s3-tools MCP Server | s3-tools MCP Server is an MCP server that connects Model Context Protocol clients directly to Amazon Simple Storage Service (AWS | Unknown | https://github.com/MCP-Mirror/sofianhamiti_mcp-server-s3 |
| RunPod MCP Server | RunPod MCP Server is an MCP server that connects Claude and other Model Context Protocol clients to the RunPod REST | Unknown | https://github.com/antonioevans/runpod-mcp-ts |
| S3 Tools | S3 Tools is an MCP server that provides tools for interacting with Amazon Web Services (AWS) Simple Storage Service (S3). | Unknown | https://github.com/sofianhamiti/mcp-server-s3 |
| Vercel Domains MCP | Vercel Domains MCP is an MCP server that enables AI assistants to query and inspect domain details registered or hosted | Unknown | https://github.com/wong2/vercel-domains-mcp |
For Kubernetes workloads, mcp-k8s-go provides a Golang-based MCP server designed specifically for cluster interactions. It is built to be extensible, allowing language models to inspect and manage containerized infrastructure without manually executing kubectl commands.
Tools like Shell MCP Server and Shell Executor provide managed execution environments. Shell MCP Server validates commands against configurable directories, shell types, and execution timeouts, while Shell Executor focuses on safely triggering commands on remote host machines.
Both s3-tools MCP Server and S3 Tools are configured to provide object storage capabilities. They expose tools that allow AI assistants to query buckets, inspect objects, and manage files stored on AWS S3 within supported MCP client interfaces.
RunPod MCP Server lets AI clients interface with the RunPod REST API to allocate and monitor GPU instances. For ephemeral developer environments, Shipyard lets agents pull logs, run tests, compare branches, and start or stop environments directly.
Yes. Sherlock Domains provides tools to search, buy, and configure domains through Sherlock, while Vercel Domains MCP enables assistants to query domain configurations registered under your Vercel organization.