TRIGGERcmd is an MCP server that enables AI models to run commands remotely on your computers through the TRIGGERcmd cloud automation platform. The server connects LLM clients directly to your configured TRIGGERcmd account, allowing system administrators, homelab enthusiasts, and software developers to manage remote operating systems and execute shell commands using natural language. It operates via standard input and output (stdio), offering baseline tools to discover and run commands across connected machines, as well as dynamic tool generation for commands that include specific descriptions. Users configure command triggers within their TRIGGERcmd dashboard, defining target hosts, native shell scripts, and optional parameter definitions. Once connected to an MCP-compatible assistant, the server inspects available triggers and exposes them as native functions that accept runtime arguments. This setup allows users to launch local applications, perform system maintenance routines, trigger home automation scripts, or query status on remote Windows, Mac, or Linux systems without manually opening terminal sessions or exposing insecure inbound network ports.
Category: Cloud & Infrastructure
Tags: automation, command-line, remote-execution, shell
triggercmd-mcp-darwin-arm64 or triggercmd-mcp-windows-amd64.exe), or pull the container image triggercmd-mcp:latest. For macOS and Linux binaries, run chmod +x <binary-name>. 2. Configure authentication by setting the TRIGGERCMD_TOKEN environment variable or by creating a token file at ~/.TRIGGERcmdData/token.tkn. 3. Open your client configuration file, such as claude_desktop_config.json for Claude Desktop or mcp.json for VS Code. 4. Add the server entry under mcpServers pointing to your executable path: json { "mcpServers": { "triggercmd": { "command": "C:\\Users\\johndoe\\Downloads\\triggercmd-mcp-windows-amd64.exe" } } } 5. Restart your MCP client to initialize the server and load your available computer commands.Part of MCP Servers
TRIGGERcmd is an MCP server that links Model Context Protocol clients with the TRIGGERcmd remote execution service. It allows AI models to run shell commands on your Windows, Mac, or Linux computers remotely by executing actions preconfigured in your account.
The server provides base tools named list_commands and run_command to inspect and trigger scripts across computers. It also generates dynamic tools following the pattern run_computer_command for triggers that have their mcpToolDescription field filled out in the TRIGGERcmd account dashboard.
TRIGGERcmd works with any client supporting standard input and output MCP servers, including Claude Desktop and VS Code. You can configure it using a locally downloaded binary or by running the server inside a Docker container with your authentication token mounted.
You can authenticate by setting the TRIGGERCMD_TOKEN environment variable in your MCP client configuration or operating system. Alternatively, the server can read an existing token from the token.tkn file located in your ~/.TRIGGERcmdData directory on the host computer.