MCP File Finder

What is MCP File Finder?

MCP File Finder is an MCP (Model Context Protocol) server that enables searching for files on the filesystem by matching a fragment of the file path. It is intended for developers or any users who need a simple, local HTTP API to locate files on a machine.

How to use MCP File Finder?

Install Python 3.8 or later, clone the repository, install dependencies with pip install -r requirements.txt, then run the server with python mcp_server.py. The server starts on http://127.0.0.1:5000. Send GET requests to the /find endpoint with query (file path fragment) and dir (starting directory) parameters.

Key features of MCP File Finder

• Searches files by any fragment of the file path.
• Returns results as JSON with name, full path, size, and creation date.
• Simple HTTP API: one endpoint (/find) with two parameters.
• Lightweight – runs locally with no external services required.
• Open-source and easy to modify.

Use cases of MCP File Finder

• Quickly locate a file on your development machine using only a partial path.
• Automate file discovery in scripts or other tools that consume JSON.
• Integrate file search capabilities into an MCP‑compatible client or agent.

FAQ from MCP File Finder

Does MCP File Finder require any external databases or services?

No, it is a self‑contained Python server that scans the local filesystem on each request.

What Python version is needed?

Python 3.8 or higher is required.

Where does the server listen by default?

It binds to http://127.0.0.1:5000.

What transport protocol does it use?

The server uses plain HTTP. No authentication mechanisms are mentioned.

Are there any known limits on file system depth or number of results?

The README does not specify limits; the server will search recursively from the given dir path.