mcp-cli-exec MCP Server
A CLI command execution server that enables running shell commands with structured output, providing detailed execution results including stdout, stderr, exit code, and execution duration.
jakenuts
Tools
cli-exec-raw
Execute a raw CLI command and return structured output
cli-exec
Execute one or more CLI commands in a specific working directory
README
mcp-cli-exec MCP Server
A powerful CLI command execution MCP server that enables running shell commands with structured output. This package focuses specifically on command execution functionality, differentiating it from other MCP CLI tools.
Features
Tools
cli-exec-raw
Execute a raw CLI command and return structured output
- Takes a command string and optional timeout (default: 5 minutes)
- Returns detailed execution results including stdout, stderr, exit code
- Handles errors gracefully with structured error responses
cli-exec
Execute one or more CLI commands in a specific working directory
- Supports single commands, && chained commands, or array of commands
- All commands execute in the specified working directory
- Returns detailed results for each command:
- Success/failure status
- Exit code
- stdout and stderr (ANSI codes stripped)
- Execution duration
- Working directory
- Stops on first command failure
- Optional timeout per command (default: 5 minutes)
Note: Due to execution context limitations, each command runs independently. Directory changes (cd) within commands do not affect subsequent commands. All commands execute in the initially specified working directory.
Output Format
Commands return structured results including:
- Success/failure status
- Exit code
- stdout and stderr (with ANSI codes stripped)
- Execution duration
- Working directory
- Detailed error information if applicable
Example Usage
cli-exec-raw
Simple command execution:
{
"command": "echo Hello World"
}
With timeout:
{
"command": "long-running-script.sh",
"timeout": 300000
}
cli-exec
Single command in specific directory:
{
"workingDirectory": "/path/to/project",
"commands": "npm install"
}
Multiple commands (all run in the same working directory):
{
"workingDirectory": "C:\\project",
"commands": [
"dir /b",
"npm run build"
]
}
Installation
Optionally install from npm:
npm install -g mcp-cli-exec
# or with pnpm
pnpm add -g mcp-cli-exec
Or just use npx in your configuration
For Cline VSCode Extension
Add to %APPDATA%/Code - Insiders/User/globalStorage/rooveterinaryinc.roo-cline/settings/cline_mcp_settings.json
:
{
"mcpServers": {
"mcp-cli-exec": {
"command": "npx",
"args": ["-y", "mcp-cli-exec"]
}
}
}
For Claude Desktop
Add to the appropriate config file:
Windows: %APPDATA%/Claude/claude_desktop_config.json
MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"mcp-cli-exec": {
"command": "npx",
"args": ["-y", "mcp-cli-exec"]
}
}
}
Special Windows Configuration
If you encounter the ENOENT spawn npx issue on Windows, use this alternative configuration that specifies the full paths:
{
"mcpServers": {
"mcp-cli-exec": {
"command": "C:\\Users\\jim\\AppData\\Roaming\\nvm\\v22.1.0\\node.exe",
"args": [
"C:\\Users\\jim\\AppData\\Roaming\\npm\\node_modules\\npm\\bin\\npx-cli.js",
"-y",
"mcp-cli-exec"
]
}
}
}
Development
Install dependencies:
pnpm install
Build the server:
pnpm run build
For development with auto-rebuild:
pnpm run watch
Debugging
Since MCP servers communicate over stdio, debugging can be challenging. The MCP Inspector provides helpful debugging tools:
pnpm run inspector
This will provide a URL to access the inspector in your browser, where you can:
- View all MCP messages
- Inspect request/response payloads
- Test tools interactively
- Monitor server state
Error Handling
The server includes comprehensive error handling:
- Input validation for all tool parameters
- Structured error responses
- Command timeout handling
- Working directory validation
- ANSI code stripping for clean output
Technical Details
- Built with TypeScript and the MCP SDK
- Uses execa for reliable command execution
- Default command timeout: 5 minutes
- Supports Windows and Unix-like systems (use appropriate commands for your OS, e.g., 'dir' vs 'ls')
- Executes commands sequentially, stopping on first failure
- Each command runs independently in the specified working directory
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
MCP Package Docs Server
Facilitates LLMs to efficiently access and fetch structured documentation for packages in Go, Python, and NPM, enhancing software development with multi-language support and performance optimization.
Claude Code MCP
An implementation of Claude Code as a Model Context Protocol server that enables using Claude's software engineering capabilities (code generation, editing, reviewing, and file operations) through the standardized MCP interface.
@kazuph/mcp-taskmanager
Model Context Protocol server for Task Management. This allows Claude Desktop (or any MCP client) to manage and execute tasks in a queue-based system.
Linear MCP Server
Enables interaction with Linear's API for managing issues, teams, and projects programmatically through the Model Context Protocol.
mermaid-mcp-server
A Model Context Protocol (MCP) server that converts Mermaid diagrams to PNG images.
Jira-Context-MCP
MCP server to provide Jira Tickets information to AI coding agents like Cursor

Linear MCP Server
A Model Context Protocol server that integrates with Linear's issue tracking system, allowing LLMs to create, update, search, and comment on Linear issues through natural language interactions.

Sequential Thinking MCP Server
This server facilitates structured problem-solving by breaking down complex issues into sequential steps, supporting revisions, and enabling multiple solution paths through full MCP integration.