Shell Executor
Execute shell commands safely on a remote server.
Shell Executor MCP Server
A Model Context Protocol (MCP) server that provides safe shell command execution capabilities for AI agents and tools like VS Code Copilot.
Features
- Safe Command Execution: Whitelist-based command filtering
- Comprehensive Tool Set: Support for file operations, text processing, development tools, and system utilities
- MCP Compatible: Full Model Context Protocol support
- Easy Integration: Simple JSON-RPC interface over stdio
Installation
Option 1: Direct Usage
git clone https://github.com/yourusername/shell-executor-mcp.git
cd shell-executor-mcp
pip install -e .
Option 2: Install from PyPI (when published)
pip install shell-executor-mcp
VS Code Integration
1. Global Configuration
Add to your VS Code settings.json or MCP configuration:
{
"mcpServers": {
"shell-executor": {
"command": "python",
"args": ["/path/to/shell-executor-mcp/src/mcp_server.py"],
"env": {
"MCP_API_TOKEN": "your-secret-token"
}
}
}
}
2. Workspace Configuration
Create .vscode/mcp_servers.json in your workspace:
{
"shell-executor": {
"command": "python",
"args": ["${workspaceFolder}/path/to/mcp_server.py"],
"env": {
"MCP_API_TOKEN": "your-secret-token"
}
}
}
3. Using with VS Code Copilot
Once configured, Copilot can discover and use the shell executor through MCP:
@shell-executor execute ls -la
@shell-executor execute git status
@shell-executor execute cargo build
Manual Testing
Start the server:
python src/mcp_server.py
Send test requests:
- Initialize:
{"jsonrpc": "2.0", "id": 1, "method": "initialize", "params": {"protocolVersion": "2024-11-05", "capabilities": {}, "clientInfo": {"name": "test", "version": "1.0.0"}}}
- List tools:
{"jsonrpc": "2.0", "id": 2, "method": "tools/list", "params": {}}
- Execute command:
{"jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": {"name": "execute_command", "arguments": {"command": "echo", "args": ["Hello World"]}}}
Supported Commands
File System
ls,cat,head,tail,find,tree,wc,du,dffile,stat,pwd
Text Processing
grep,sed,awk,sort,uniq,cut,tr
Development
git,npm,yarn,pip,python,node,cargo,makerustc,go,java,javac
System Info
ps,uptime,whoami,id,uname,which
Network
curl,wget,ping,dig,nslookup
Archives
tar,zip,unzip,gzip,gunzip
Utilities
echo,date,cal,bc,expr,basename,dirname- Hash tools:
md5,sha256sum,shasum
Security
- Commands are filtered through a whitelist
- 30-second execution timeout
- No destructive operations (rm, mv, cp) allowed
- No privilege escalation (sudo, su) allowed
- Environment variables can be controlled
Configuration
Set environment variables:
MCP_API_TOKEN: Authentication token (default: "MYSECRET")
License
MIT License - see LICENSE file for details.
Máy chủ liên quan
Scout Monitoring MCP
nhà tài trợPut performance and error data directly in the hands of your AI assistant.
Alpha Vantage MCP Server
nhà tài trợAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
MetaMCP
A self-hostable middleware to manage all your MCPs through a GUI and a local proxy, supporting multiple clients and workspaces.
Railagent
Provides structured workflows and tools for AI agents working with software development projects, with a specialized focus on Rails applications.
DALL-E Image Generator
Generate images using OpenAI's DALL-E API.
MCP Stdio Server
An MCP server using stdio transport, offering file system access, a calculator, and a code review tool. Requires Node.js.
R.A.P.I.D.
A local MCP server providing powerful code analysis and search capabilities for software projects.
cesium-mcp
AI-powered CesiumJS 3D globe control — 43 tools for camera, entities, layers, animation, and interaction via MCP protocol. Also available as a remote server via Streamable HTTP.
MCP Software Engineer Server
An MCP server that enables an AI to function as a full-stack software engineer with complete development capabilities.
TechDebtMCP
MCP server for analyzing and managing technical debt in codebases via the Model Context Protocol
tachibot-mcp
Stop AI Hallucinations Before They Start Run models from OpenAI, Google, Anthropic, xAI, Perplexity, and OpenRouter in parallel. They check each other's work, debate solutions, and catch errors before you see them.
Aluvia
The Aluvia MCP server exposes browser session management, geo-targeting, and account operations as Model Context Protocol tools for AI agents.