MCP Script Runner
Execute developer-defined bash scripts in a Dockerized environment for coding agents.
MCP Script Runner
A Model Context Protocol (MCP) server that provides coding agents with a generic interface to execute developer-defined bash scripts within a Dockerized environment.
🚀 Quick Start
Option 1: Docker (Recommended)
# Clone and run with Docker
git clone <repository-url>
cd devmcp
docker compose up --build -d
# Check logs
docker compose logs -f
Option 2: Local Installation
# Clone repository
git clone <repository-url>
cd devmcp
# Install dependencies
pip install -r requirements.txt
# Run the server
python -m mcp_script_runner.server
📋 Overview
The MCP Script Runner Server enables AI agents to:
- Execute predefined bash scripts with configurable arguments
- Manage working directories for different project contexts
- Get script information including descriptions and available arguments
- List available scripts dynamically
- Handle script timeouts and error conditions gracefully
🐳 Docker Support
The project includes full Docker support for containerized execution:
- 🔧 Ready-to-use containers with all dependencies
- 🛡️ Isolated execution environment
- 📦 Easy deployment with Docker Compose
- 🔍 Debug capabilities with interactive shell access
See DOCKER.md for complete Docker usage guide.
Quick Docker Commands
# Start the MCP server
docker compose up --build -d
# Interactive development shell
docker compose --profile debug up shell
# Test script execution
docker compose exec mcp-script-runner bash scripts/hello.sh
🛠️ Installation
Prerequisites
- Python 3.11+
- Docker (for containerized execution)
- Bash-compatible shell
Local Setup
# Install Python dependencies
pip install -r requirements.txt
# Verify installation
python -c "from src.mcp_script_runner.server import main; print('✅ Installation OK')"
Docker Setup
# Build container
docker build -t mcp-script-runner .
# Or use Docker Compose
docker compose up --build
⚙️ Configuration
Configuration File: .mcp-config.json
{
"working_directory": ".",
"scripts": {
"hello": {
"path": "scripts/hello.sh",
"description": "Simple hello world script",
"arguments": [],
"timeout": 30
},
"list_files": {
"path": "scripts/list_files.sh",
"description": "List files in directory with options",
"arguments": ["directory", "options"],
"timeout": 10
}
}
}
Script Directory Structure
project/
├── .mcp-config.json
├── scripts/
│ ├── hello.sh
│ ├── list_files.sh
│ └── system_info.sh
└── src/
└── mcp_script_runner/
🔧 Available MCP Tools
| Tool | Description | Arguments |
|---|---|---|
run_script | Execute a configured script | script_name, arguments[] |
list_scripts | List all available scripts | None |
get_script_info | Get script details | script_name |
get_working_directory | Get current working directory | None |
set_working_directory | Set working directory | path |
reload_config | Reload configuration file | None |
Example Tool Usage
{
"tool": "run_script",
"arguments": {
"script_name": "hello",
"arguments": []
}
}
🏃♂️ Running the Server
Local Execution
# Start MCP server (listens on stdio)
python -m mcp_script_runner.server
# Or with explicit path
PYTHONPATH=src python -m mcp_script_runner.server
Docker Execution
# Background service
docker compose up -d
# Interactive mode
docker compose run --rm mcp-script-runner
# Debug shell
docker compose --profile debug up shell
📝 Example Scripts
Basic Hello Script (scripts/hello.sh)
#!/bin/bash
echo "Hello from MCP Script Runner!"
echo "Current directory: $(pwd)"
echo "Script arguments: $@"
echo "Date: $(date)"
File Listing Script (scripts/list_files.sh)
#!/bin/bash
DIRECTORY=${1:-.}
OPTIONS=${2:-"-la"}
echo "Listing files in: $DIRECTORY"
ls $OPTIONS "$DIRECTORY"
System Info Script (scripts/system_info.sh)
#!/bin/bash
echo "=== System Information ==="
echo "OS: $(uname -s)"
echo "Kernel: $(uname -r)"
echo "Architecture: $(uname -m)"
echo "Uptime: $(uptime)"
echo "Disk Usage:"
df -h
🧪 Testing
Unit Tests
# Run tests locally
python -m pytest tests/
# Run tests in Docker
docker compose run --rm mcp-script-runner python -m pytest tests/
Manual Testing
# Test script execution
python -c "
import asyncio
from src.mcp_script_runner.executor import ScriptExecutor
from src.mcp_script_runner.config import ConfigManager
async def test():
cm = ConfigManager()
ex = ScriptExecutor(cm)
result = await ex.execute_script('hello')
print(f'Exit code: {result.exit_code}')
print(result.stdout)
asyncio.run(test())
"
🔐 Security Considerations
- 🛡️ Containerized execution isolates script execution
- 👤 Non-root user inside containers (
mcpuser) - 📁 Limited file access through volume mounts
- ⏱️ Script timeouts prevent runaway processes
- 🚫 No shell injection - arguments passed safely
🎯 Use Cases
Development Automation
- Build and test commands
- Code generation scripts
- Development environment setup
System Administration
- System monitoring scripts
- Backup and maintenance tasks
- Configuration management
CI/CD Integration
- Deployment scripts
- Environment validation
- Automated testing workflows
Project Management
- Task automation
- Report generation
- Resource management
🐛 Troubleshooting
Common Issues
MCP Server Won't Start
# Check Python path
export PYTHONPATH=src
# Verify dependencies
pip install -r requirements.txt
# Check configuration
python -c "from src.mcp_script_runner.config import ConfigManager; cm = ConfigManager(); print('Config OK')"
Script Execution Fails
# Check script permissions
chmod +x scripts/*.sh
# Test script directly
bash scripts/hello.sh
# Check Docker logs
docker compose logs mcp-script-runner
Docker Issues
# Rebuild container
docker compose up --build
# Check container status
docker compose ps
# Interactive debugging
docker compose run --rm mcp-script-runner bash
Debug Mode
# Local debug
PYTHONPATH=src python -c "
import logging
logging.basicConfig(level=logging.DEBUG)
from mcp_script_runner.server import main
import asyncio
asyncio.run(main())
"
# Docker debug
docker compose --profile debug up shell
📚 Development
Project Structure
devmcp/
├── 📄 README.md # This file
├── 🐳 DOCKER.md # Docker usage guide
├── 📋 TASKS.md # Development tasks
├── ⚙️ .mcp-config.json # Configuration
├── 🐳 Dockerfile # Container definition
├── 🐳 docker-compose.yml # Container orchestration
├── 📦 requirements.txt # Python dependencies
├── 📦 pyproject.toml # Python project config
├── 🔧 scripts/ # Example scripts
├── 🐍 src/mcp_script_runner/ # Python source code
└── 🧪 tests/ # Unit tests
Adding New Scripts
- Create script in
scripts/directory - Make executable:
chmod +x scripts/myscript.sh - Add to
.mcp-config.json:
{
"scripts": {
"myscript": {
"path": "scripts/myscript.sh",
"description": "My custom script",
"arguments": ["arg1", "arg2"],
"timeout": 30
}
}
}
- Reload configuration: Use
reload_configtool
Contributing
- Fork the repository
- Create feature branch
- Add tests for new functionality
- Test with Docker:
docker compose up --build - Submit pull request
🚀 Deployment
Production Deployment
# Using Docker Compose
docker compose up -d
# Using Docker Swarm
docker stack deploy -c docker-compose.yml mcp-stack
# Using Kubernetes
kubectl apply -f k8s/
Integration with MCP Clients
Claude Desktop Configuration
{
"mcpServers": {
"script-runner": {
"command": "docker",
"args": ["compose", "-f", "/path/to/devmcp/docker-compose.yml", "run", "--rm", "mcp-script-runner"]
}
}
}
📄 License
MIT License - see LICENSE file for details.
🤝 Support
- 📖 Documentation: See DOCKER.md for Docker usage
- 🐛 Issues: Create GitHub issue
- 💬 Discussions: GitHub Discussions
- 📧 Contact: See repository contributors
Ready to get started?
🐳 Docker users: docker compose up --build
🐍 Local users: pip install -r requirements.txt && python -m mcp_script_runner.server
Máy chủ liên quan
Alpha Vantage MCP Server
nhà tài trợAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
Ghost MCP
An MCP server for the Ghost blogging platform with Server-Sent Events (SSE) transport support.
PHP MCP Server for Laravel
A Laravel wrapper for the php-mcp/server library to expose Laravel applications as MCP servers.
VSCode MCP Server
A VSCode extension that acts as an MCP server, providing access to diagnostic tools and debug session management.
GoPeak
The most comprehensive MCP server for Godot Engine. 95+ tools for scene management, GDScript LSP, DAP debugging, screenshot capture, input injection, and CC0 asset library.
Mastra/mcp
Client implementation for Mastra, providing seamless integration with MCP-compatible AI models and tools.
Coding Prompt Engineer MCP Server
Rewrites coding prompts for optimal results with AI IDEs like Cursor AI, powered by Claude by Anthropic.
PowerShell MCP Server
Automate Windows PowerShell tasks using Python. Execute scripts, manage the clipboard, and capture terminal output programmatically.
ng-diagram MCP Server
Provides ng-diagram documentation and API lookup tools for AI coding assistants.
Moralis
Interact with the Moralis Web3 API to access blockchain data and services.
AgentLux MCP Server
Installable MCP server for AgentLux marketplace, identity, creator, services, social, and Base/x402 commerce flows.