Ecs MCP Server

🚀 360 AI Cloud Disk MCP

This project provides a Model Context Protocol (MCP) access service for the 360 AI Cloud Disk. It enables AI models to directly operate the cloud disk via the MCP protocol, offering comprehensive cloud disk file management capabilities.

🚀 Quick Start

The 360 AI Cloud Disk's MCP (Model Context Protocol) service allows various AI models (such as large language models) to interact with the 360 AI Cloud Disk through the standard MCP protocol. In this way, AI models can assist users in managing cloud disk files, significantly enhancing the intelligence and convenience of file management.

✨ Features

This MCP service offers a variety of operations for interacting with the 360 AI Cloud Disk, including:

• 📁 Browse file lists - View the contents of cloud disk directories.
• 🔍 Search for files - Search for cloud disk files by keywords.
• ⬆️ Upload files - Upload files to the 360 Cloud Disk.
• ⬇️ Download files - Obtain download links for cloud disk files and support direct downloads.
• 🎬 Download videos - Download videos to the cloud disk via URLs, supporting batch downloads and real - time progress monitoring.
• 💾 Save files - Save files to the cloud disk via URLs or text content.
• 📂 Create directories - Create new folders in the cloud disk.
• ✏️ Rename files - Modify the names of cloud disk files or folders.
• 🚚 Move files - Move files to other locations.
• 🔗 Share files - Generate share links for specified files.
• 🔑 Get user information - Retrieve user personal information.

📦 Installation

Configuration in Cursor

Stdio Access Method

Add the following configuration to the ~/.cursor/mcp.json file to connect to the 360 AI Cloud Disk MCP service:

{
  "mcpServers": {
    "360-mcp-server-disk": {
      "command": "npx",
      "args": [
        "-y",
        "@aicloud360/mcp-server-disk"
      ],
      "env": {
        "API_KEY": "_xxxxxxxxx"
      }
    }
  }
}

Streamable HTTP Access Method

If you prefer to access via HTTP, use the following configuration:

{
  "mcpServers": {
    "mcp-server-disk-http": {
      "url": "https://mcp.yunpan.com/mcp?api_key=_xxxxxxxxx"
    }
  }
}

Characteristics of the Streamable HTTP access method:

• No need to install the Node.js environment.
• No need to download and run locally.
• Authenticate by passing the API_KEY as a URL parameter.
• Suitable for integration scenarios that require HTTP interfaces.

SSE Access Method

If you want to access via SSE (Server - Sent Events), use the following configuration:

{
  "mcpServers": {
    "mcp-server-disk-sse": {
      "url": "https://mcp.yunpan.com/sse?api_key=_xxxxxxxxx"
    }
  }
}

Characteristics of the SSE access method:

• Based on the HTTP long - connection server - push technology.
• Stronger real - time performance, suitable for scenarios that require immediate responses.
• One - way communication, where the server pushes data to the client.
• No need to install an additional environment, natively supported by browsers.
• Authenticate by passing the API_KEY as a URL parameter.

🔐 Authentication Configuration

To use the 360 AI Cloud Disk MCP service, you need the following authentication information:

• API_KEY: The 360 AI Cloud Disk API key, in the format of a string starting with "yunpan_".

You can obtain the API_KEY in the following way:

• Refer to Quick Access to the MCP Server.

Advantages of the 360 AI Cloud Disk Open Platform

The 360 AI Cloud Disk Open Platform provides diversified product capabilities and one - stop file services:

• One - click account association: No need to register a new account. You can associate your existing 360 AI Cloud Disk account with one click, achieving a seamless login experience of "multiple platforms, one account".
• Support for MCP protocol access: Supports Stdio/SSH/Streamable HTTP/SSE protocols, and can be easily accessed through the MCP Client.
• Rich interface capabilities: Provides APIs for file upload, download, search, creation, renaming, moving, sharing, etc., meeting the needs of different scenarios.

Visit the 360 AI Cloud Disk Open Platform official website for more detailed information and the latest development documentation.

💻 Usage Examples

File Upload (file-upload-stdio) - Only supports Stdio access method

Upload local files to a specified path on the 360 AI Cloud Disk.

Parameters:

• filePaths: The full path of local files (required, can be an array of strings containing multiple files).
• uploadPath: The target directory on the cloud disk for uploading, defaulting to the root directory /.

Example:

{
  "filePaths": ["/Users/username/Documents/Report.docx", "/Users/username/Documents/Data.xlsx"],
  "uploadPath": "/Work Files"
}

Single file upload example:

{
  "filePaths": "/Users/username/Desktop/Test Document.pdf",
  "uploadPath": "/Documents"
}

File Download (file-download-stdio) - Only supports Stdio access method

Obtain the download link of a specified file on the cloud disk and support direct file download.

Parameters:

• nid: The unique identifier ID of the file, which can be obtained through the file list or search (required).
• auto: Whether to directly download the file, defaulting to true.
• downloadDir: The specified download directory, which must have read and write permissions, defaulting to the .mcp-downloads folder under the user's home directory.

Example of only getting the download link:

{
  "nid": "12345678",
  "auto": false
}

Example of downloading to a specified directory:

{
  "nid": "12345678",
  "auto": true,
  "downloadDir": "/Users/username/Downloads"
}

File List Query (file-list)

Get the list of files and folders in a specified path on the 360 AI Cloud Disk.

Parameters:

• path: The path to query, defaulting to the root directory /.
• page: The page number, defaulting to 0.
• page_size: The number of items displayed per page, defaulting to 50.

Example:

{
  "path": "/Documents",
  "page": 1,
  "page_size": 20
}

File Search (file-search)

Search for files on the 360 AI Cloud Disk by keywords.

Parameters:

• key: The search keyword (required).
• file_category: The file type (-1: all, 0: others, 1: images, 2: documents, 3: music, 4: videos), defaulting to -1.
• page: The page number, defaulting to 1.
• page_size: The number of items displayed per page, defaulting to 20.

Example:

{
  "key": "Report",
  "file_category": 2,
  "page": 1
}

File Save (file-save)

Save files to the cloud disk via URLs or text content.

Parameters:

• url: The file download address (either url or content must be provided).
• content: The file content, supporting Markdown format (either url or content must be provided).
• upload_path: The storage path on the cloud disk, which must start with /, defaulting to /From: mcp_server/.

Example of saving via URL:

{
  "url": "https://example.com/sample.pdf",
  "upload_path": "/Documents/Downloads/"
}

Example of saving via text content:

{
  "content": "# Title\nThis is a Markdown - formatted text content",
  "upload_path": "/Notes/"
}

Video Download (video-download)

Download videos to the cloud disk via URLs, supporting batch downloads and real - time progress monitoring. This operation may take a long time. It is recommended to set a longer timeout for the client (recommended to be more than 300 seconds).

Parameters:

• urls: Video URLs, separated by the English vertical bar '|' for multiple URLs (required).

Single video download example:

{
  "urls": "https://example.com/video.mp4"
}

Batch video download example:

{
  "urls": "https://example.com/video1.mp4|https://example.com/video2.mp4|https://example.com/video3.mp4"
}

Functional features:

• 🎯 Batch download: Supports downloading multiple video URLs simultaneously.
• 📊 Real - time progress: Provides detailed download progress monitoring, including task status distribution.
• 🔄 Automatic polling: Automatically polls the task status until completion, eliminating the need for manual queries.
• 📁 Cloud disk storage: Downloaded videos are directly saved to the cloud disk, providing cloud disk file links.
• ⚡ Status tracking: Tracks task status in real - time (pending/ downloading/ download successful/ upload successful/ failed).
• 🔗 Convenient access: Provides cloud disk file links after completion for easy direct access.

Explanation of return results:

• Successfully downloaded videos will display information such as the cloud disk file path, file size, and access link.
• Failed videos will display the specific failure reason.
• Supports structured data return for easy program processing.

Create Folder (make-dir)

Create a new folder in the 360 AI Cloud Disk.

Parameters:

• fname: The folder path, for example: /New Folder/ (required).

Example:

{
  "fname": "/Work Files/Project A/"
}

File Share (file-share)

Generate a share link for a specified file.

Parameters:

• paths: The paths of files to be shared, separated by the vertical bar (|) for multiple file paths (required).

Example:

{
  "paths": "/Documents/Report.docx|/Documents/Data.xlsx"
}

Move File (file-move)

Move files or folders in the 360 AI Cloud Disk to a new location.

Parameters:

• src_name: The original path of the file, separated by the vertical bar for multiple paths (required).
• new_name: The target path (required).

Example:

{
  "src_name": "/Documents/Report.docx|/Documents/Data.xlsx",
  "new_name": "/Archived Folders/"
}

Rename File (file-rename)

Rename files or folders in the 360 AI Cloud Disk.

Parameters:

• src_name: The original path name, for example: /Folder/Old File Name.txt (required).
• new_name: The new name, for example: New File Name.txt (required).

Example:

{
  "src_name": "/Documents/Draft.docx",
  "new_name": "Final Report.docx"
}

User Information (user-info)

Get the personal information of a 360 AI Cloud Disk user.

Parameters:

• None

🧠 AI Application Scenarios

Through the 360 AI Cloud Disk MCP access, AI can help users achieve the following scenarios:

• Intelligent file organization: AI can analyze the content of user files and automatically categorize and organize them.
• Intelligent document retrieval: Use natural language descriptions to find documents in the cloud disk.
• Automatic file backup: Provide intelligent backup suggestions based on user habits.
• File content analysis: Analyze document content and provide summaries or insights.
• Dialog - based file operations: Users can manage cloud disk files through dialogues.
• File upload and sharing: Users can save files to the cloud disk through dialogues and generate file share links, making it easy to share files with others.
• Save online resources: Users can provide URLs and let AI help save online resources to the cloud disk.
• Create and save file content: AI can create document content according to user needs and directly save it to the cloud disk.
• Download cloud disk files: Users can download files from the cloud disk to the local through dialogues.
• Download video resources: Users can provide video URLs and let AI help download videos to the cloud disk, supporting batch downloads and progress monitoring.

🔑 Keywords

• 360 AI Cloud Disk
• mcp
• modelcontextprotocol
• ai assistant
• file management
• video download
• batch download
• sse
• streamable http

📄 License

Apache - 2.0