raycblai/s3-select-mcp

S3 Select MCP Server

This MCP server provides tools to query S3 objects using S3 Select with natural language queries.

Features

• List files in an S3 bucket
• Translate natural language queries to S3 Select SQL queries
• Execute S3 Select queries on specific files
• Search across multiple files with a single natural language query

Setup

1. Install the required dependencies:

pip install -r requirements.txt

2. Configure AWS credentials:

Make sure you have the bedrockuser profile configured in your AWS credentials file.

3. Start the MCP server:

python s3_select_mcp_server.py

4. Register the MCP server with Amazon Q CLI:

q mcp add s3-select http://localhost:8080

Usage

Once the server is running and registered with Amazon Q, you can use the following tools:

• s3-select___list_s3_files: List all files in the configured S3 bucket
• s3-select___translate_to_s3_select_query: Convert natural language to S3 Select SQL
• s3-select___execute_s3_select_query: Run an S3 Select query on a specific file
• s3-select___natural_language_s3_select: Search across files using natural language

Example Queries

• "Find all records where age is greater than 30"
• "Show me sales data from January 2023"
• "List customers from California with purchases over $1000"

Configuration

The server is configured to use:

• AWS Profile: bedrockuser
• AWS Region: us-west-2
• S3 Bucket: amazon-q-280792572112-artifacts
• S3 Prefix: s3-select-data

To change these settings, modify the constants at the top of the s3_select_mcp_server.py file.