Redshift MCP Server

Example Use Cases

• 1

  Run safe, read-only SQL queries against an Amazon Redshift cluster using the Redshift Data API and return results to an LLM.

• 2

  Explore and discover database structure—listing schemas, tables, and retrieving table DDL and metadata—to answer schema-related queries.

• 3

  Diagnose performance, inspect execution plans, detect lock contention, and monitor workload patterns for Redshift troubleshooting.

Description

This project implements a Model Context Protocol (MCP) server to bridge LLMs or AI assistants with Amazon Redshift. It uses the AWS Redshift Data API and AWS Secrets Manager for secure connections and exposes MCP tools for schema discovery, ad-hoc read-only queries, table inspection, query performance diagnosis, workload monitoring, and cluster health checks. The server standardizes access via the MCP spec so compatible clients (e.g., Claude, Cursor) can integrate with Redshift in a controlled, auditable way. It is targeted at developers, data engineers, and analysts who want to add natural-language or agent-driven access to Redshift data and diagnostics.