Cloud Storage MCP Server

A Model Context Protocol (MCP) server for Google Cloud Storage that enables interactions with Google Cloud Storage buckets and files.

Features

• List Cloud Storage buckets in a project
• Get details of a specific bucket
• List files in a bucket
• Get details of a specific file
• Upload files to a bucket
• Download files from a bucket
• Delete files from a bucket

Setup

1. Install dependencies:

   npm install
   

   Copy

2. Build the project:

   npm run build
   

   Copy

3. Configure Claude Desktop: Add the following to your claude_desktop_config.json:

   "cloudstorage-mcp": {
     "command": "node",
     "args": [
       "/path/to/cloudstorage-mcp/build/index.js"
     ],
     "env": {
       "GOOGLE_CLOUD_PROJECTS": "project-id1,project-id2"
     }
   }
   

   Copy

   Replace the path in args with the actual path to index.js.

   Define a comma-separated list of project IDs in GOOGLE_CLOUD_PROJECTS. Example: google-project-id1,google-project-id2 The first listed project is the default.

   The application expects to find .json credential file(s) in the keys folder for each project. Example: keys/google-project-id1.json

   Ensure the relevant cloud service account has appropriate permission to interact with Cloud Storage, e.g. Storage Admin or lesser permission(s).

Available Tools

• listBuckets: List all Cloud Storage buckets in a project
• getBucket: Get details of a specific Cloud Storage bucket
• listFiles: List files in a Cloud Storage bucket
• getFile: Get details of a specific file in a Cloud Storage bucket
• uploadFile: Upload a file to a Cloud Storage bucket
• downloadFile: Download a file from a Cloud Storage bucket
• deleteFile: Delete a file from a Cloud Storage bucket

Example Usage in Claude Desktop

Here are examples of how to use each tool in Claude Desktop:

List Buckets

List all buckets in my Google Cloud project.

Get Files in a Bucket

Show me all files in the backup-data bucket.

Get File Details

Get details of the file reports/monthly_report.pdf in the data-analysis bucket.

Development

# Watch mode
npm run dev