MCP Workspace Server

MCP Workspace Server

Provides secure, sandboxed file system access for AI assistants to read, write, and manage project files with controlled command execution capabilities, all confined to a designated workspace directory.

Category
Visit Server

README

๐Ÿ† ULTIMATE MCP Workspace Server

License: MIT TypeScript Node.js Tools Categories

The Most Comprehensive MCP Server Ever Created - 36 professional tools across 15+ categories for development, DevOps, data processing, and automation. Transform Claude Desktop into the ultimate development environment!

๐ŸŒŸ Ultimate Features

  • ๐Ÿ† 36 Comprehensive Tools: The most complete MCP server available
  • ๐Ÿ”’ Enterprise Security: Military-grade sandboxing and path validation
  • ๐Ÿš€ 15+ Categories: File ops, Git, Docker, cloud storage, databases, and more
  • ๐ŸŒ Multi-Platform: Windows, macOS, and Linux support
  • โ˜๏ธ Multi-Cloud: AWS S3, Google Cloud, Azure integration
  • ๐Ÿณ DevOps Ready: Docker management, package managers, CI/CD tools
  • ๐Ÿ“Š Data Processing: CSV, JSON, databases, web scraping, image processing
  • ๐Ÿ” Security Suite: Encryption, hashing, secure key generation
  • โฐ Automation: Task scheduling, notifications, webhooks
  • ๐ŸŽจ Code Quality: Formatting, linting, analysis tools
  • โœ… Production Ready: 109+ passing tests, comprehensive error handling

๐Ÿ“‹ Table of Contents

๐Ÿš€ Installation

Prerequisites

  • Node.js v18 or higher (Download)
  • npm (comes with Node.js)
  • An MCP-compatible client (Claude Desktop, Cline, etc.)

Setup Steps

  1. Clone the repository

    git clone https://github.com/ShayYeffet/ultimate_mcp_server.git
    cd ultimate_mcp_server
    
  2. Install dependencies

    npm install
    
  3. Build the project

    npm run build
    
  4. Verify installation

    npm test
    

You should see all 109 tests passing! โœ…

โšก Quick Start

For Claude Desktop Users

  1. Find your Claude config file:

    • Windows: %APPDATA%\Claude\claude_desktop_config.json
    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    • Linux: ~/.config/Claude/claude_desktop_config.json
  2. Create or edit the file with this configuration:

    {
      "mcpServers": {
        "workspace": {
          "command": "node",
          "args": ["/ABSOLUTE/PATH/TO/ultimate_mcp_server/dist/index.js"],
          "env": {
            "MCP_WORKSPACE_ROOT": "/path/to/your/project",
            "MCP_ALLOWED_COMMANDS": "npm,git,node",
            "MCP_LOG_LEVEL": "info"
          }
        }
      }
    }
    
  3. Restart Claude Desktop

  4. Test it! Ask Claude:

    "List all files in my workspace"
    "Create a new file called test.txt with 'Hello World'"
    

โš™๏ธ Configuration Guide

๐Ÿ”ง What You MUST Change

When setting up the MCP server, you must customize these values in your config file:

1. Path to the Server (Required)

"args": ["/ABSOLUTE/PATH/TO/ultimate_mcp_server/dist/index.js"]

Replace with:

  • Windows: "C:\\Users\\YourName\\path\\to\\ultimate_mcp_server\\dist\\index.js"
  • macOS/Linux: "/home/username/path/to/ultimate_mcp_server/dist/index.js"

How to find it:

# In the ultimate_mcp_server directory, run:
pwd  # macOS/Linux
cd   # Windows

2. Workspace Root Directory (Required)

"MCP_WORKSPACE_ROOT": "/path/to/your/project"

This is THE MOST IMPORTANT setting! This directory is where the AI can read/write files.

Examples:

  • Your web project: "C:\\Users\\YourName\\projects\\my-website"
  • Your app: "/home/username/projects/my-app"
  • A test folder: "C:\\Users\\YourName\\ai-workspace"

โš ๏ธ Security Note: The AI can ONLY access files inside this directory. Choose carefully!

3. Allowed Commands (Recommended)

"MCP_ALLOWED_COMMANDS": "npm,git,node"

Customize based on your needs:

  • Web development: "npm,git,node,yarn"
  • Python projects: "python,pip,git"
  • No commands: "" (empty string - disables command execution)
  • Multiple commands: "npm,git,python,node,cargo,go"

โš ๏ธ Security Note: Only list commands you trust the AI to run!

๐Ÿ“ Optional Configuration

Read-Only Mode

Prevent ALL write operations (useful for analysis only):

"MCP_READ_ONLY": "true"

Logging Level

Control how much logging you see:

"MCP_LOG_LEVEL": "debug"  // Options: debug, info, warn, error

Command Timeout

Set maximum time for commands (in milliseconds):

"MCP_COMMAND_TIMEOUT": "600000"  // 10 minutes

๐Ÿ“‹ Complete Configuration Examples

Example 1: Web Development Project

{
  "mcpServers": {
    "workspace": {
      "command": "node",
      "args": ["C:\\Users\\John\\ultimate_mcp_server\\dist\\index.js"],
      "env": {
        "MCP_WORKSPACE_ROOT": "C:\\Users\\John\\projects\\my-react-app",
        "MCP_ALLOWED_COMMANDS": "npm,git,node,yarn",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Example 2: Python Data Science Project

{
  "mcpServers": {
    "workspace": {
      "command": "node",
      "args": ["/home/jane/ultimate_mcp_server/dist/index.js"],
      "env": {
        "MCP_WORKSPACE_ROOT": "/home/jane/projects/data-analysis",
        "MCP_ALLOWED_COMMANDS": "python,pip,git,jupyter",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Example 3: Read-Only Code Review

{
  "mcpServers": {
    "workspace": {
      "command": "node",
      "args": ["/Users/alex/ultimate_mcp_server/dist/index.js"],
      "env": {
        "MCP_WORKSPACE_ROOT": "/Users/alex/code-to-review",
        "MCP_ALLOWED_COMMANDS": "",
        "MCP_READ_ONLY": "true",
        "MCP_LOG_LEVEL": "warn"
      }
    }
  }
}

Example 4: Multiple Workspaces

You can configure multiple MCP servers for different projects:

{
  "mcpServers": {
    "project-a": {
      "command": "node",
      "args": ["C:\\ultimate_mcp_server\\dist\\index.js"],
      "env": {
        "MCP_WORKSPACE_ROOT": "C:\\projects\\project-a",
        "MCP_ALLOWED_COMMANDS": "npm,git"
      }
    },
    "project-b": {
      "command": "node",
      "args": ["C:\\ultimate_mcp_server\\dist\\index.js"],
      "env": {
        "MCP_WORKSPACE_ROOT": "C:\\projects\\project-b",
        "MCP_ALLOWED_COMMANDS": "python,pip"
      }
    }
  }
}

๐Ÿ› ๏ธ Ultimate Tool Arsenal - 36 Tools

๐Ÿ“ Core File Operations (7 Tools)

  • list_files - List directory contents with filtering and sorting
  • read_file - Read file contents with encoding support
  • write_file - Create or update files atomically
  • delete_file - Delete files or directories safely
  • create_folder - Create directories with parent creation
  • apply_patch - Apply unified diff patches to files
  • run_command - Execute shell commands securely

๐Ÿ” Advanced File Operations (5 Tools)

  • search_files - Grep-like text search across files with patterns
  • find_files - Find files by name, pattern, size, date, extension
  • copy_file - Copy files and directories recursively
  • move_file - Move or rename files and folders
  • get_file_info - Detailed file metadata, checksums, permissions

๐Ÿ“ฆ Archive & Compression (2 Tools)

  • compress_files - Create ZIP/TAR/GZ archives with compression
  • extract_archive - Extract compressed archives safely

๐ŸŒ Network Operations (1 Tool)

  • http_request - Full HTTP client with all methods, headers, timeouts

๐Ÿ”ง Git Operations (4 Tools)

  • git_status - Repository status, changes, branch information
  • git_diff - Show differences between commits, branches, files
  • git_log - View commit history with filters and formatting
  • git_branch - Branch management, creation, deletion, listing

๐Ÿ’ป System Operations (3 Tools)

  • system_info - CPU, memory, disk, network, OS information
  • list_processes - Show running processes with resource usage
  • kill_process - Terminate processes by PID or name

๐Ÿ—„๏ธ Database Operations (1 Tool)

  • database_query - Execute SQL queries on SQLite databases

๐Ÿ–ผ๏ธ Image Processing (1 Tool)

  • image_process - Resize, crop, rotate, convert image formats

๐Ÿ“„ PDF Manipulation (1 Tool)

  • pdf_manipulate - Extract text, get info, merge, split PDFs

๐Ÿ” Encryption & Security (1 Tool)

  • encrypt_decrypt - AES-256 encryption, hashing (SHA-256/512, MD5), key generation

โฐ Task Scheduling (1 Tool)

  • schedule_task - Schedule commands with cron/Windows Task Scheduler

๐Ÿ“ข Notifications (1 Tool)

  • send_notification - System notifications, webhooks, email alerts

๐Ÿ“ Text Processing (1 Tool)

  • text_process - Analyze, transform, extract patterns, compare, generate text

๐Ÿ“Š Data Processing (2 Tools)

  • csv_process - Read, write, transform, analyze, filter, merge, split CSV files
  • json_process - Parse, validate, transform, merge, extract, minify, prettify JSON

๐Ÿ•ท๏ธ Web Scraping (1 Tool)

  • web_scrape - Fetch HTML, extract elements, links, images, text, metadata

๐Ÿณ Docker Management (1 Tool)

  • docker_manage - Manage containers, images, networks, volumes, logs, exec

โ˜๏ธ Cloud Storage (1 Tool)

  • cloud_storage - Upload, download, list, delete files on AWS S3, GCP, Azure

๐Ÿ“ฆ Package Management (1 Tool)

  • package_manager - Install, uninstall, update packages with npm, yarn, pip, composer, gem, cargo, go

๐ŸŽจ Code Formatting (1 Tool)

  • code_format - Format, lint, analyze, fix code with prettier, eslint, black, etc.

๐Ÿ’ก Example Commands

Development Workflow

"Search for all TODO comments in TypeScript files"
"Show git status and list all modified files"
"Format all JavaScript files with prettier"
"Create a ZIP archive of the src folder"
"Run npm install and build the project"

System Administration

"Show system information including CPU and memory usage"
"List all processes using more than 100MB of memory"
"Schedule a daily backup at 2 AM"
"Send a notification when the deployment completes"
"Kill the process using port 3000"

Data & Content Processing

"Query the database for all users created this month"
"Resize all images in this folder to 800x600"
"Extract text from this PDF and save to a file"
"Analyze this CSV file and show statistics"
"Scrape product data from this website"

Cloud & DevOps

"Upload build artifacts to AWS S3"
"List all running Docker containers"
"Install dependencies with npm and pip"
"Make a POST request to the API with this data"
"Encrypt sensitive files with AES-256"

๐Ÿ”’ Security

Sandboxing

All file operations are strictly sandboxed to the MCP_WORKSPACE_ROOT directory:

โœ… Allowed:

read_file("src/index.ts")
write_file("output/data.json", content)

โŒ Blocked:

read_file("../../../etc/passwd")
read_file("/etc/passwd")
write_file("C:\\Windows\\System32\\file.txt", content)

Protection Mechanisms

  1. Path Resolution: All paths resolved to absolute paths and validated
  2. Boundary Checking: Resolved paths must start with workspace root
  3. Traversal Detection: ../ sequences that escape workspace are blocked
  4. Absolute Path Blocking: Absolute paths outside workspace rejected
  5. Symbolic Link Resolution: Symlinks resolved and validated
  6. Command Allowlist: Only explicitly allowed commands can execute
  7. Argument Safety: Command arguments passed as arrays (no shell injection)

Best Practices

  • โœ… Use a dedicated workspace directory for AI operations
  • โœ… Only allow commands you trust
  • โœ… Use read-only mode for code review/analysis
  • โœ… Regularly review what the AI is doing
  • โœ… Keep the workspace separate from system directories
  • โŒ Don't set workspace root to / or C:\
  • โŒ Don't allow dangerous commands like rm, del, format

๐Ÿงช Testing

The project includes comprehensive test coverage:

# Run all tests
npm test

# Run tests in watch mode
npm run test:watch

# Run with coverage
npm run test:coverage

Test Structure

  • Unit Tests: Test individual functions and modules (26 tests)
  • Property-Based Tests: Use fast-check to verify properties across random inputs (83 tests)
  • Integration Tests: Test full MCP protocol communication

All 109 tests validate the correctness properties defined in the design specification.

๐Ÿ› Troubleshooting

Server Won't Start

Problem: Claude/Cline shows connection errors

Solutions:

  1. Verify Node.js is installed: node --version (should be 18+)
  2. Check the server path in config is correct and absolute
  3. Ensure the project is built: npm run build
  4. Check that dist/index.js exists

Security Errors

Problem: "Path outside workspace" errors

Solutions:

  1. Ensure MCP_WORKSPACE_ROOT is an absolute path
  2. Use relative paths in requests (no leading /)
  3. Avoid ../ to navigate outside workspace
  4. Check for symbolic links pointing outside workspace

Command Not Allowed

Problem: "Command not allowed" errors

Solutions:

  1. Add command to MCP_ALLOWED_COMMANDS: "npm,git,python"
  2. Use only the command name, not full path
  3. Ensure commands are comma-separated without spaces

Command Timeout

Problem: Commands killed before completing

Solutions:

  1. Increase MCP_COMMAND_TIMEOUT (value in milliseconds)
  2. For long builds: "MCP_COMMAND_TIMEOUT": "1800000" (30 min)
  3. Check if command is actually hanging

File Not Found

Problem: AI can't find files that exist

Solutions:

  1. Verify MCP_WORKSPACE_ROOT points to correct directory
  2. Use relative paths from workspace root
  3. Check file actually exists: ls or dir in workspace

JSON Parse Error (Claude Desktop)

Problem: "Could not load app settings" or "invalid JSON"

Solutions:

  1. Validate JSON syntax at jsonlint.com
  2. Ensure all paths use double backslashes on Windows: C:\\Users\\...
  3. Check for trailing commas (not allowed in JSON)
  4. Verify file encoding is UTF-8 without BOM

๐Ÿ“š Environment Variables Reference

Variable Type Default Description
MCP_WORKSPACE_ROOT string process.cwd() REQUIRED: Absolute path to workspace directory
MCP_ALLOWED_COMMANDS string "" Comma-separated list of allowed commands
MCP_READ_ONLY boolean false Disable all write operations
MCP_LOG_LEVEL string "info" Logging level: debug, info, warn, error
MCP_COMMAND_TIMEOUT number 300000 Command timeout in milliseconds (5 min default)

๐Ÿค Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Development Setup

  1. Fork the repository
  2. Clone your fork
  3. Install dependencies: npm install
  4. Make your changes
  5. Run tests: npm test
  6. Build: npm run build
  7. Submit a PR

Guidelines

  • Ensure all tests pass
  • Add tests for new features
  • Follow TypeScript strict mode
  • Update documentation as needed

๐Ÿ“„ License

This project is licensed under the MIT License - see the LICENSE file for details.

๐Ÿ™ Acknowledgments

  • Built following the Model Context Protocol specification
  • Tested with fast-check for property-based testing
  • Inspired by the need for secure AI-filesystem interaction

๐Ÿ“ž Support


โญ If you find this project useful, please consider giving it a star on GitHub!

Recommended Servers

playwright-mcp

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.

Official
Featured
TypeScript
Magic Component Platform (MCP)

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.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
E2B

E2B

Using MCP to run code via e2b.

Official
Featured