redshift-guide

Amazon Redshift Guide

Redshift is NOT PostgreSQL (read first)

Redshift speaks PostgreSQL's wire protocol and shares much of its surface syntax, so LLMs assume PostgreSQL behavior carries over — it frequently does not. Divergences span system tables (pg_catalog is incomplete), DDL (no indexes, no sequences), functions (string_agg, SUBSTR on tables, leader-node-only functions), types (a text column becomes VARCHAR(256)), and comparison semantics (trailing blanks, unenforced constraints). Assume divergence and verify against the reference below — do not answer from PostgreSQL habit. Common PostgreSQL→Redshift divergences are in references/redshift-sql-syntax.md.

This guide ships with this server. Run the SQL it recommends, SHOW commands and system views included, through execute_query, one statement per call. The list_* tools discover clusters, databases, schemas, tables and columns, and review_cluster runs a cluster review. To load a reference this guide cites, call retrieve_skill_file with the path exactly as cited. This server runs no AWS CLI command: give the user the ones this guide shows, or run them with another tool you have. The database user this server connects as sees only its own rows (the queries it generated) in the system views, with no error, unless it is a superuser or has sys:monitor or SYSLOG ACCESS UNRESTRICTED; some views refuse it with permission denied instead → references/redshift-auth-authorization.md.

STEP 0: Serverless or Provisioned?

Establish this before answering — APIs, system tables, and capabilities differ. Take it from the question when it says which one, else from the type of the target's list_clusters entry; ask only when neither settles it. SELECT version() does not identify it.

• Serverless — identified by a workgroup (and namespace). Data API calls take --workgroup-name; the user says "workgroup"/"Serverless".
• Provisioned — identified by a cluster. Data API calls take --cluster-identifier; the user says "cluster".