Terminal MCP Server

What is Terminal MCP Server?

Terminal MCP Server is a Model Context Protocol (MCP) server that allows executing commands on local or remote hosts. It provides a simple yet powerful interface for AI models and other applications to execute system commands, either on the local machine or on remote hosts via SSH.

How to use Terminal MCP Server?

Install manually via git clone, npm install, and npm run build, or automatically via Smithery using npx -y @smithery/cli install @weidwonder/terminal-mcp-server --client claude. Start the server in stdio mode with npm start or in SSE mode with npm run start:sse. Configure the server URL in your AI assistant’s MCP settings file (e.g., Claude Desktop, Cline, or Roo Code).

Key features of Terminal MCP Server

• Execute commands directly on the local machine
• Execute commands on remote hosts via SSH
• Persistent sessions that reuse the same terminal environment (default 20 minutes)
• Set custom environment variables per command
• Connect via stdio or SSE (Server-Sent Events)

Use cases of Terminal MCP Server

• AI assistants running system commands on the user’s local machine
• Remotely executing commands on a server via SSH from an AI chat interface
• Maintaining a persistent terminal session across multiple related command calls

FAQ from Terminal MCP Server

What parameters does the execute_command tool accept?

The tool accepts a required command string, optional host and username for remote execution, optional session name (defaults to "default"), and optional env object for environment variables.

How do I connect to a remote host?

Specify the host and username parameters in the execute_command tool. SSH key-based authentication must be set up in advance; the default key path is ~/.ssh/id_rsa.

How long do sessions persist?

Sessions automatically close after 20 minutes of inactivity. Using the same session name across commands reuses the same terminal environment within that window.

What transport modes are supported?

The server supports stdio (local connection) and SSE (Server-Sent Events for remote HTTP connections). SSE mode can be customized with --port, --host, and --endpoint flags.

Where does data live and what are the security requirements?

Commands run in the context of the user who started the server. For local execution, no additional setup is needed. For remote execution, SSH key-based authentication must be configured on the target host.