AWS S3 access tool for MCP Clients 🪣⚙️

This tool is an AWS S3 bucket utility built for MCP Server, allowing seamless interaction with S3 storage. It lets LLMs list available buckets, retrieve stored objects, and filter for CSV files, with built-in async support using aioboto3. Developers can integrate either locally or via Docker. The tool is optimized for automation, making S3 access smoother and more scalable.

https://github.com/user-attachments/assets/976dce46-6ee3-4c1a-b5eb-8ca1575df099

The tool supports environment-based bucket selection, meaning you can restrict access to specific buckets using an .env file.

Key Features

• ✅ List available S3 buckets
• ✅ Fetch objects from specific or all buckets
• ✅ Filter and retrieve only CSV files
• ✅ Read the content of a CSV file from S3

Requirements

Ensure you have the following dependencies installed:

pip install aioboto3 mcp[cli] python-dotenv

Environment Setup

AWS_ACCESS_KEY_ID=your_access_key
AWS_SECRET_ACCESS_KEY=your_secret_key
AWS_REGION=your_region
S3_BUCKETS=bucket1,bucket2  # (Optional) List of buckets to access

MCP Client Configuration

• Configuration for Cursor.ai

   {
   "mcpServers": {
       "AWS-S3-AccessTool": {
           "command": "python",
           "args": ["C:your-absolute-path-to-the-file\\main.py"]
  
   		}
   	}
   }
  

License ⚖️

This project is open-source and available under the MIT License.