Getting Started with MCP Filesystem Server: Docker Deployment Guide

Getting Started with MCP Filesystem Server: Step-by-Step Guide

This tutorial will walk you through deploying and using the MCP Filesystem Server with Docker. By the end, you'll have a fully functional filesystem server that your AI agents can use to read, write, and manage files on your local machine.

Prerequisites

• Docker installed on your system (Docker Engine 20.10+)
• MCP-compatible client such as Claude Desktop, VS Code with GitHub Copilot, or Cursor
• Node.js 18+ (optional, for npm-based installation)

Method 1: Docker (Recommended)

The fastest way to get started is with the official Docker image:

# Pull the latest image
docker pull mcp/filesystem:latest

# Run with access to your current directory
docker run -i --rm \
  -v /path/to/allowed/dir:/data \
  mcp/filesystem:latest \
  /data

The -i flag is essential — MCP servers communicate via stdio, so the container must run interactively. Mount the directories you want the AI agent to access as volumes.

Method 2: Docker Compose

For a more permanent setup, create a docker-compose.yml file:

services:
  mcp-filesystem-server:
    image: mcp/filesystem:latest
    restart: unless-stopped
    volumes:
      - ./projects:/data
      - ./config:/config
    stdin_open: true
    tty: true

Then start it:

docker compose up -d

Method 3: Direct npm Install

If you prefer running without Docker:

# Install globally
npm install -g @modelcontextprotocol/server-filesystem

# Run with specific directory access
mcp-server-filesystem /path/to/projects /path/to/config

Configuring with AI Clients

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "filesystem": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/me/projects:/data",
        "mcp/filesystem:latest",
        "/data"
      ]
    }
  }
}

VS Code (GitHub Copilot)

Configure in VS Code settings under github.copilot.mcpServers:

{
  "github.copilot.mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-filesystem",
        "/path/to/projects"
      ]
    }
  }
}

What You Can Do

Once configured, your AI agent can:

• Read files: "Read the contents of package.json"
• Write files: "Create a new React component at src/components/Header.tsx"
• Search: "Find all TODO comments in the project"
• Manage directories: "Create a new folder for API routes"
• Get metadata: "What's the size of the largest file in this directory?"

Security Tips

• Only mount directories that the AI agent needs — never the entire filesystem
• Use read-only mounts when the agent only needs to read files
• On multi-user systems, create separate volumes for each user
• Review access patterns with MCP Roots for dynamic, audit-friendly control

Troubleshooting

• Permission denied: Ensure the Docker container has read/write permissions on mounted volumes
• Agent can't find files: Verify the paths passed to the server match the mounted volume paths
• Connection refused: Make sure stdio is properly piped when using docker run -i

Conclusion

The MCP Filesystem Server is one of the simplest yet most powerful tools in the MCP ecosystem. With just a single Docker command, you can give your AI agents the ability to work directly with your files — saving time, reducing copy-paste errors, and enabling fully autonomous development workflows.