MCP Shell Server
by mkusaka
MCP Shell Server is a bridge that allows AI agents to safely execute shell commands using the Model Context Protocol (MCP). It functions as a server to execute shell commands.
Last updated: N/A
MCP Shell Server
A server that uses the Model Context Protocol (MCP) to execute shell commands. It functions as a bridge that allows AI agents to safely execute shell commands.
Features
- Execute shell commands (single-line and multi-line support)
- Support for various shells (bash, zsh, fish, powershell, cmd, etc.)
- Detailed error handling and logging
- MCP Inspector compatible
Installation
From npm (as a user)
# Using npm
npm install -g @mkusaka/mcp-shell-server
# Using yarn
yarn global add @mkusaka/mcp-shell-server
# Using pnpm
pnpm add -g @mkusaka/mcp-shell-server
From source (for development)
# Clone the repository
git clone https://github.com/mkusaka/mcp-shell-server.git
cd mcp-shell-server
# Install dependencies
pnpm install
# Build the project
pnpm build
MCP Configuration
Cursor Configuration
Add the following to your Cursor configuration file (~/.cursor/config.json
):
{
"mcpServers": {
"shell": {
"command": "npx",
"args": ["-y", "@mkusaka/mcp-shell-server"]
}
}
}
Rule Configuration
Add the following to your Cursor rules file:
You have tools at your disposal to solve the coding task. Follow these rules regarding tool calls:
1. ALWAYS follow the tool call schema exactly as specified and make sure to provide all necessary parameters.
2. The conversation may reference tools that are no longer available. NEVER call tools that are not explicitly provided.
3. **NEVER refer to tool names when speaking to the USER.** For example, instead of saying 'I need to use the shell_exec tool to execute your command', just say 'I will execute your command'.
4. Only calls tools when they are necessary. If the USER's task is general or you already know the answer, just respond without calling tools.
5. Before calling each tool, first explain to the USER why you are calling it.
Usage
Direct Execution
node dist/index.js
# or as an executable
./dist/index.js
Development Mode
pnpm dev
Testing with MCP Inspector
pnpm inspect
Command Line Arguments
-s, --shell <shell> Specify the path to the shell to use
-h, --help Display help message
-V, --version Display version information
Tool Reference
shell_exec
Executes commands in the specified shell.
Parameters:
command
(string, required): The shell command to execute
Resource Reference
The server provides the following system information as resources:
hostname
Returns the hostname of the system.
URI: hostname://
platform
Returns the operating system platform.
URI: platform://
shell
Returns the shell path being used by the server.
URI: shell://
username
Returns the current username.
URI: username://
system-info
Returns comprehensive system information in JSON format, including:
- hostname
- platform
- shell
- username
- CPU count
- Total memory
- Free memory
- System uptime
Usage Examples
Basic Command Execution
{
"name": "shell_exec",
"parameters": {
"command": "echo Hello, World!"
}
}
Multi-line Command (Heredoc) Execution
{
"name": "shell_exec",
"parameters": {
"command": "cat << EOF | grep 'example'\nThis is an example text.\nAnother line without the keyword.\nEOF"
}
}
Development
Project Structure
src/
├── index.ts # Main entry point
└── shell-server/
├── index.ts # Shell server implementation
└── lib/
└── logger.ts # Logging configuration
Logging
Logs are written to the mcp-shell.log
file.
License
MIT