MCPRisingofficial publisher

Filesystem MCP

MCP : Standardises the connection between an AI application and systems that expose tools or data · requires tout client MCP

Reference MCP server for reading and writing files within authorized folders.

Basic and safe if you strictly limit the exposed folders.

GitHub · 91,545 ★

Overall score

7.8 /10

Security: one caveat, see the card

Utility

8.0

Reliability

10.0

Security

One caveat

Ease

5 min · Comfortable

Security reflects where data goes, requested permissions and incidents on record at the verification date. It is not a guarantee of zero risk: check the deployment mode and the connections you enable.

What it does, concretely

Managing files from an AI assistant usually means manually copying and pasting content between your editor and the chat window. Filesystem MCP removes that friction by giving your AI client direct access to folders you define in advance.

The server is installed via npx and declared in your MCP client configuration. You pass the list of allowed folders as arguments: the AI can then read, write, and navigate within those folders without going outside that perimeter.

This tool is aimed at developers and technical users who want to automate file tasks on their local machine from Claude, Cursor, or VS Code, without going through a graphical interface.

It is not designed for production use or shared environments: the official repository explicitly states that it is a reference implementation intended for educational purposes, not a production-ready solution.

What it offers

·File reading within declared folders, accessible directly from the AI client.
·File writing and editing within the authorized perimeter, without manual intervention.
·Directory tree navigation within exposed folders to list and explore files.
·Configurable access control: only paths passed as arguments at startup are accessible.
·Compatible with any MCP client, including Claude Desktop, Cursor, and VS Code.
·No permanent installation required thanks to npx, with no global dependency to manage.

Our tips

Pass only the strictly necessary folders as arguments: a dedicated project folder rather than your entire home directory.
On Windows, wrap the npx command with cmd /c to avoid launch errors from certain MCP clients.
Declare multiple folders by adding them one after another in the args array of your JSON configuration.
Test first with an empty or test folder to verify that the access perimeter is properly restricted before pointing to sensitive files.

Watch out

The official repository warns that this server is a reference implementation for educational purposes, not a production-ready solution: assess your own security requirements before any serious use.
An overly broad folder exposed by mistake (such as the full home directory) gives the AI write access to critical files with no additional safeguard.

In practice

Price: Free
Commercial use: Commercial use: yes
Available on: claude-code, cursor, vscode
Host: tout client MCP
Licence: MIT

Security and data

Rating: Security: one caveat, see the card
Your data: Data stays local
Maintenance: Active
Last release: 25 October 2026

Links and repository checked on 25 October 2026 · added on 1 November 2024

How to install Filesystem MCP

1. Make sure Node.js is installed on your machine (npx is included). 2. Open your MCP client configuration file (for example the Claude Desktop configuration file). 3. Add a "filesystem" block in the "mcpServers" section, specifying the npx command and the allowed folder(s) as arguments. 4. On Windows only, replace "command": "npx" with "command": "cmd" and prepend "/c" and "npx" to the args array. 5. Restrict the paths passed as arguments to only the necessary folders, never your full home directory. 6. Restart your MCP client for the configuration to take effect.

npx -y @modelcontextprotocol/server-filesystem /path/to/allowed/files

In the same category

Whole category Feed (memory, context) →