execute_command
Execute a command in the local system
LABSCO SUMMARY
One tool, and that tool is the shell.
execute_command takes a command, its args and a cwd, which is enough to drive a build, a test run or a version-control operation from a conversation. The surface carries no allowlist, no confirmation argument and no output cap, so what this server can do is exactly what its account can do on that machine.
What it is
A single-tool server that runs commands on the machine hosting it.
What you get
- Run a named command with its own argument list on the local system.
- Set the working directory per call through cwd, so a build or test runs where the project actually is.
Requirements
Commands execute on the machine hosting the server, with that account's permissions, and there is nothing to configure first.
Setup effort
One command — npx -y @dillip285/mcp-terminal
USE CASES
01
Running builds and tests without leaving the conversation
Point the call at the project directory and let the model run the command and read the output back.
1 tool
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
BEFORE YOU INSTALL
What to know first
Tools you get
1 tool when we connected.
Requires
npx on PATH
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the one tool definition that rides along on every turn
INSTALLATION
Connect your client
Server setup
TERMINAL
npx -y @dillip285/mcp-terminalMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"terminal": {
"command": "npx",
"args": [
"@dillip285/mcp-terminal"
]
}
}
}The project's README adds --allowed-paths /path/to/allowed/directory here, but the server never reads it: neither its source nor the published 1.0.1 build parses command-line arguments. Commands run with your account's full permissions.
