kyle-chalmers

Claude Code Redshift — Development skill for Claude Code

Development community

Connect Claude Code to Amazon Redshift: run a multi-step analysis read-only, then read the query plan to find and fix a distribution-key problem.

How to install Claude Code Redshift

This entry records only its repository, not the path inside it, so there is no exact command to give. Open kyle-chalmers/claude-code-redshift and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Code Redshift does

Connect Claude Code to Amazon Redshift: run a multi-step analysis read-only, then read the query plan to find and fix a distribution-key problem. Companion to the video.

Alternatives in Development

Investigate Performance — Investigate a Nimbalyst performance problem (freeze, lag, idle CPU, slow op, memory growth) with measured evid 1.6k ★
Fullstack Solution Template For Agentcore — Flexible Fullstack solution template for production-ready deployments of any use case on Amazon Bedrock AgentC 471 ★
Dashclaw Quality — Recurring find-and-fix quality pass over the whole DashClaw app — browser smoke (frontend-verify) + code gates 297 ★

README

Claude Code + Amazon Redshift: run the analysis, then fix the slow query

Point [Claude Code](https://www.anthropic.com/claude-code) at an Amazon Redshift warehouse and let it do two things most demos never show. First, run a real multi-step analysis on its own: write a query, read the result, ask the next question, and keep going until it has an answer. Second, when one of those queries is slow, read Redshift's own query plan, find the distribution key that is dragging it down, and write the fix for you to apply.

This repo is the companion to the video. It gives you a small synthetic warehouse (seeded on purpose with one performance problem and one accuracy trap), the connection setup, and the exact prompts, so you can reproduce the whole thing in a few minutes.

Channel: [Kyle Chalmers Data Plus AI](https://www.youtube.com/@kylechalmersdataai)

![What runs where: Claude Code and the connection on your machine, the Data API to Redshift Serverless, and what leaves for the model API](./images/architecture.png)

How it connects (two official paths)

Redshift is the fourth database in this series, and it has two first-party connection options worth knowing about. Both authenticate with IAM through the [Redshift Data API](https://docs.aws.amazon.com/redshift/latest/mgmt/data-api.html), so there is no database password sitting in a config file.

	`aws redshift-data` CLI	AWS Labs Redshift MCP server
What it is	The agent runs `aws redshift-data` commands	Structured tools the agent calls
Auth	IAM via the Data API, no password	IAM via the Data API, no password
Read vs write	Whatever the IAM role allows	Read-only by default
Surface	Any SQL or command	Six scoped tools (list + query)
Best for	Full control, writes, scripting	Safe exploration and analysis

This repo uses the [AWS Labs Redshift MCP server](https://github.com/awslabs/mcp/tree/main/src/redshift-mcp-server) for the analysis, because read-only by def