Gemini MCP File Agent (MVP)

@kevingduck

リポジトリ

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

作者

kevingduck

Gemini MCP File Agent について

Simple chat interface and server for testing Gemini MCP

#gemini #mcp#file-agent

Gemini MCP File Agent の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "gemini_mcp": {
      "command": "python",
      "args": [
        "mcp_server.py"
      ]
    }
  }
}

Gemini MCP File Agent のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Gemini MCP File Agent (MVP)?

It is a local server and chat scripts that allow Google's Gemini AI to safely read, write, and list files on your computer within a specific, controlled folder (a sandbox). It is intended for developers exploring AI file access.

How to use Gemini MCP File Agent (MVP)?

Install the Python packages, set your Google AI API key as the GOOGLE_API_KEY environment variable, then run mcp_server.py (optionally with --sandbox-dir) in one terminal and chat_with_gemini_mcp.py in another. You can then ask Gemini about files in the sandbox.

Key features of Gemini MCP File Agent (MVP)

Safe file access limited to a sandbox folder.
Command-line chat interface with Gemini.
Can read, write, and list files.
Uses a local MCP server (mcp_server.py).
Designed for local, single‑computer use.

Use cases of Gemini MCP File Agent (MVP)

Ask Gemini what files are in your work folder.
Have Gemini create a file and write content to it.
Have Gemini read the contents of an existing file.
Explore giving an AI controlled file access.

よくある質問

What is the sandbox?

The sandbox is a specific folder you choose (default: `./mcp_data_sandbox/`). The server can only touch files inside that folder for safety.

Why do I need an API key?

You need a Google AI API key from Google AI Studio to use Gemini. Set it as the environment variable `GOOGLE_API_KEY`.

Can Gemini access files outside the sandbox?

No. The `mcp_server.py` is designed to only access files inside the folder you pick as the sandbox.

How do I use a different sandbox folder?

Run `python mcp_server.py --sandbox-dir ./my_files` to use a custom folder instead of the default `./mcp_data_sandbox/`.

Is this safe for production use?

No. The README states this is a basic example and cautions you to be careful about which folder the server accesses.

「AI とエージェント」の他のコンテンツ

OpenZiti MCP Gateway

openziti

NetFoundry MCP Gateway enables secure, isolated access to Model Context Protocol (MCP) tools across distributed systems without exposing public endpoints. Zero trust gateway for MCP servers. Aggregate, filter, and secure

Gemina

Gemina Tech Ltd.

Extract, search and tag any document — invoices, receipts, contracts, forms, or your own templates. 13 tools, one sign-in (OAuth 2.1) or API key, free FileTag tier, data residency in the EU, US, Israel or Asia.

Hostinger

hostinger

Model Context Protocol (MCP) server for Hostinger API. If you don't want to install or run anything locally, connect directly to Hostinger's hosted MCP server: This opens a browser window to authorize via OAuth. Once app

148

OpenZiti / LLM-Gateway