DP-MCP Server

You are setting up and using a DP-MCP Server that connects PostgreSQL databases with MinIO storage, enriched by AI-powered capabilities. This server lets you run SQL queries, manage schemas, export data, handle object storage, and leverage intelligent AI-assisted analysis—all through the MCP protocol. It runs locally via FastMCP over HTTP and can be controlled with a CLI, a local stdio workflow, or a remote HTTP client.

How to use

You interact with the DP-MCP Server using an MCP client or the included CLI. Start the server locally, then connect your client to the MCP endpoint at http://127.0.0.1:8888/mcp/. Use the available tools to perform PostgreSQL operations, bucket and object management in MinIO, and AI-powered analysis. You can run SQL queries, inspect schemas, export data, back up database tables to object storage, and trigger AI-driven insights and natural language questions about your data. In production, you may deploy on a different host/port and enable debug or demo modes as needed.

How to install

Follow these concrete steps to install and run the DP-MCP Server in a development environment.

How to use with the MCP client

Connect your MCP client to the server URL and call the available tools to perform database, storage, and AI actions. The server exposes HTTP-based MCP endpoints and supports FastMCP transport for streaming results.

Configuration and security

Configure connections to PostgreSQL and MinIO, and set up AI models and privacy levels. Use a local .env file to define credentials and server options. The server supports multiple privacy levels for AI interactions, with local processing by models like Ollama and optional cloud models if you enable API keys.

Usage examples

Typical workflows include running a backup of a PostgreSQL table to MinIO, listing buckets and objects, exporting query results to CSV, and asking natural language questions that are translated into SQL and analyzed by AI.

CLI and development tools

Use the provided CLI to access PostgreSQL and MinIO operations directly from the command line. For development, run the server locally with the standard Python startup and verify health using the built-in endpoints.

Troubleshooting

If services won’t start, verify Docker and Python environments, check the health of PostgreSQL and MinIO, and review startup logs for connection errors or misconfigurations. Use the health checks and log monitoring commands to identify and fix issues quickly.

Notes

This server focuses on providing a comprehensive data platform experience with seamless DB to storage pipelines, robust security, and AI-enabled insights. Start in debug mode for the first run to observe startup messages and verify successful connections.

Available tools

execute_sql_query

Execute SQL queries with formatting and result limits.

list_db_tables

List all tables in a given schema.

describe_db_table

Describe the structure of a specific table.

export_table_csv

Export table data to a CSV file with optional filters.

list_minio_buckets

List all available MinIO buckets.

list_bucket_objects

List objects within a MinIO bucket with optional prefix and limit.

upload_to_minio

Upload data to a MinIO bucket as an object.

download_from_minio

Download an object from MinIO.

create_minio_bucket

Create a new MinIO bucket.

delete_minio_object

Delete a specific object from a bucket.

backup_table_to_minio

Backup a PostgreSQL table to MinIO as a data export.

ask_natural_language_query

Convert a natural language question into an SQL query and run it with AI analysis.

explain_query_with_ai

Run an SQL query and get an AI-powered explanation.

get_ai_data_insights

Generate AI-driven insights for database analysis.

analyze_table_patterns

AI analysis of data patterns and quality for a table.

generate_ai_data_report

Create AI-powered data reports across tables.

get_ai_system_status

Report AI system status and configuration.