Filesystem MCP Server (@sylphlab/filesystem-mcp)

@sylphlab

Repository Homepage

Basic information

About Filesystem MCP Server

📁 Secure, efficient MCP filesystem server - token-saving batch operations with project root confinement

#nodejs #npm #typescript #filesystem #mcp #cline #ai-agent #claude-ai #model-context-protocol #roo-code

Filesystem MCP Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "filesystem-mcp": {
      "command": "npx",
      "args": [
        "@sylphx/filesystem-mcp"
      ],
      "name": "Filesystem (npx)"
    }
  }
}

Filesystem MCP Server tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

Fetch the live tool list by running this server in a temporary sandbox using the button above.

Overview

What is Filesystem MCP Server (@sylphlab/filesystem-mcp)?

The Filesystem MCP Server provides secure, token‑optimized filesystem operations for AI agents via the Model Context Protocol (MCP). It enables agents like Claude and Cline to read, write, edit, search, and manage project files in batch, confined to a project root directory.

How to use Filesystem MCP Server (@sylphlab/filesystem-mcp)?

Install via npx, bunx, Docker, or local build, then configure it in your MCP host (e.g., Cline, VSCode) using the provided JSON snippets. The server uses its current working directory as the security boundary. Once configured, agents can invoke tools such as read_content, write_content, edit_file, and others.

Key features of Filesystem MCP Server (@sylphlab/filesystem-mcp)

Project root confinement prevents path traversal
Batch processing (10+ files per request) reduces token usage
Direct API calls eliminate shell spawning overhead
Zod schema validation for all tool arguments
14 tools: file, directory, management, and permission operations
Per‑item success/failure reporting in batch results

Use cases of Filesystem MCP Server (@sylphlab/filesystem-mcp)

AI agents reading and editing multiple project files in one request
Code assistants refactoring across a codebase with batch search‑replace
Automating project setup, file generation, and directory creation
Securely managing permissions and ownership within a project root

Frequently asked questions

How does project root confinement work?

All operations are restricted to the current working directory (`cwd`) at launch. Path traversal attempts are prevented, so agents cannot access files outside that directory.

What batch operations are supported?

Most tools support batch processing, including `read_content`, `write_content`, `edit_file`, `search_files`, `replace_content`, `stat_items`, `create_directories`, `delete_items`, `move_items`, `copy_items`, `chmod_items`, and `chown_items`.

What are the runtime requirements?

The server runs on Node.js or Bun, uses TypeScript in strict mode, and can be installed via npm or Docker. For development, `pnpm` is the package manager.

How does this server compare to using shell commands?

It offers better security (root confinement), higher token efficiency (batching), lower latency (direct API vs. shell spawn), and detailed per‑item error reporting.

Where is data stored and how is it transmitted?

Data resides on the local filesystem within the project root. Communication follows the MCP standard over stdio; no external network calls are made for filesystem operations.