Add database planning presets

CHANGELOG.md

@@ -6,6 +6,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Added
+- Warehouse presets and guides for Redshift, Postgres, SQL Server, Azure SQL, MySQL, DuckDB,
+  MotherDuck, and Athena.
+
 ## [1.1.0] — 2026-05-18
 
 ### Added

README.md

@@ -75,7 +75,8 @@ uv tool install dbt-spec-kit
 dbt-specify --version
 ```
 
-Supported warehouse presets: `snowflake`, `databricks`, `trino`, and `bigquery`.
+Supported warehouse presets: `snowflake`, `databricks`, `trino`, `bigquery`, `redshift`,
+`postgres`, `sqlserver`, `azure-sql`, `mysql`, `duckdb`, `motherduck`, and `athena`.
 
 ## What init adds
 
@@ -145,6 +146,14 @@ Use `dbt-specify ci` when the lifecycle and dbt artifact checks should block a P
 - [Databricks guide](docs/warehouse-guides/databricks.md)
 - [Trino guide](docs/warehouse-guides/trino.md)
 - [BigQuery guide](docs/warehouse-guides/bigquery.md)
+- [Redshift guide](docs/warehouse-guides/redshift.md)
+- [Postgres guide](docs/warehouse-guides/postgres.md)
+- [SQL Server guide](docs/warehouse-guides/sqlserver.md)
+- [Azure SQL guide](docs/warehouse-guides/azure-sql.md)
+- [MySQL guide](docs/warehouse-guides/mysql.md)
+- [DuckDB guide](docs/warehouse-guides/duckdb.md)
+- [MotherDuck guide](docs/warehouse-guides/motherduck.md)
+- [Athena guide](docs/warehouse-guides/athena.md)
 
 ## OSS project
 

docs/getting-started.md

@@ -42,7 +42,8 @@ This creates:
 - `CLAUDE.md` — the agent orientation file (or `CLAUDE.md.dbt-specify-suggested` if you already have one)
 - `specs/` — empty directory for your first spec
 
-Supported warehouse presets: `snowflake`, `databricks`, `trino`, and `bigquery`.
+Supported warehouse presets: `snowflake`, `databricks`, `trino`, `bigquery`, `redshift`,
+`postgres`, `sqlserver`, `azure-sql`, `mysql`, `duckdb`, `motherduck`, and `athena`.
 
 ## Compose with dbt-labs/dbt-agent-skills
 
@@ -102,5 +103,5 @@ dbt-specify doctor
 - Use the [team onboarding playbook](team-onboarding-playbook.md) when presenting the workflow to an analytics engineering team.
 - Read [methodology.md](methodology.md) for the full four-phase loop.
 - Read [skills-and-sub-agents.md](skills-and-sub-agents.md) for dbt Labs skills composition and safe sub-agent delegation.
-- Read [warehouse-guides/snowflake.md](warehouse-guides/snowflake.md) or [databricks.md](warehouse-guides/databricks.md) for your warehouse's preset.
+- Read the [warehouse guides](warehouse-guides/snowflake.md) for preset-specific planning rules.
 - See [`examples/jaffle-shop-staging-overhaul/`](../examples/jaffle-shop-staging-overhaul/) for a complete worked example.

docs/warehouse-guides/athena.md

@@ -0,0 +1,26 @@
+# Athena guide
+
+Use this preset when dbt targets Amazon Athena and S3-backed tables.
+
+## What the preset adds
+
+- S3 table and file-layout planning
+- partitioning and partition-projection decisions
+- Glue catalog and Lake Formation/IAM governance review
+- workgroup, output location, and scan-cost guardrails
+- Iceberg versus Hive-style table format decisions
+
+## Good fit
+
+- lakehouse tables queried through Athena
+- Iceberg or external Hive-style datasets
+- teams that need scan-cost and S3 layout evidence in PRs
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse athena
+```
+
+dbt-spec-kit does not query Athena. dbt, the adapter, AWS credentials, and workgroup configuration
+remain outside this toolkit.

docs/warehouse-guides/azure-sql.md

@@ -0,0 +1,25 @@
+# Azure SQL guide
+
+Use this preset when dbt targets Azure SQL Database or Azure SQL Managed Instance.
+
+## What the preset adds
+
+- service tier, elastic pool, CPU, IO, and log-pressure planning
+- Query Store evidence expectations for performance-sensitive work
+- index, columnstore, and incremental load guardrails
+- Microsoft Entra, role, RLS, dynamic masking, and view-boundary review
+- firewall, private endpoint, and deployment-window callouts
+
+## Good fit
+
+- Azure-native enterprise teams
+- shared elastic pool environments
+- governed marts using Azure SQL serving layers
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse azure-sql
+```
+
+This is a planning preset. It does not configure Azure credentials, networking, or dbt profiles.

docs/warehouse-guides/bigquery.md

@@ -3,6 +3,12 @@
 The BigQuery preset adds planning requirements for partitioning, clustering, materialization, cost,
 policy tags, authorized views, and BI Engine expectations.
 
+Use it with:
+
+```bash
+dbt-specify init analytics --warehouse bigquery
+```
+
 ## Partitioning
 
 Large fact-like tables should partition by the date or timestamp column that bounds routine

docs/warehouse-guides/duckdb.md

@@ -0,0 +1,26 @@
+# DuckDB guide
+
+Use this preset when dbt targets DuckDB for local analytics, CI fixtures, or file-based workflows.
+
+## What the preset adds
+
+- local file and object-store path review
+- memory and single-process limits
+- DuckDB extension planning
+- Parquet staging and export expectations
+- governance handling for local files and CI artifacts
+
+## Good fit
+
+- local development examples
+- lightweight analytics projects
+- CI smoke tests with deterministic data files
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse duckdb
+```
+
+This preset is useful for AI-agent demos because it makes local file dependencies and reproducibility
+explicit.

docs/warehouse-guides/motherduck.md

@@ -0,0 +1,26 @@
+# MotherDuck guide
+
+Use this preset when dbt targets MotherDuck or a hybrid DuckDB/MotherDuck workflow.
+
+## What the preset adds
+
+- local versus cloud execution boundary review
+- database, share, role, and collaborator planning
+- file and object-store reproducibility checks
+- cost/quota guardrails
+- sensitive data movement review
+
+## Good fit
+
+- collaborative DuckDB-style analytics
+- shared demo and lightweight team projects
+- local-to-cloud workflows where data movement must be explicit
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse motherduck
+```
+
+The preset does not configure MotherDuck authentication. It only adds planning and governance
+guidance.

docs/warehouse-guides/mysql.md

@@ -0,0 +1,25 @@
+# MySQL guide
+
+Use this preset when dbt reads from or writes to MySQL and operational safety matters.
+
+## What the preset adds
+
+- primary versus replica versus analytics database review
+- index and `EXPLAIN` expectations for heavy joins
+- incremental watermark, batch size, and lock planning
+- engine, charset, collation, and timezone assumptions
+- grants and PII access review
+
+## Good fit
+
+- operational-source staging projects
+- MySQL replicas used for analytics
+- teams that need to prevent AI agents from overloading OLTP databases
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse mysql
+```
+
+dbt-spec-kit does not run MySQL queries. dbt and the selected adapter own execution.

docs/warehouse-guides/postgres.md

@@ -0,0 +1,25 @@
+# Postgres guide
+
+Use this preset when dbt targets Postgres or uses Postgres as an analytics-safe serving database.
+
+## What the preset adds
+
+- OLTP safety checks for primary versus replica versus analytics database
+- index and materialization planning
+- lock and transaction impact review
+- schema/grant/access planning
+- extension and non-portable SQL callouts
+
+## Good fit
+
+- smaller analytics projects on Postgres
+- product analytics replicas
+- local or self-managed dbt deployments where lock safety matters
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse postgres
+```
+
+The preset does not connect to Postgres. dbt and its adapter handle database execution.

docs/warehouse-guides/redshift.md

@@ -0,0 +1,26 @@
+# Redshift guide
+
+Use this preset when dbt runs on Amazon Redshift or Redshift is the analytics serving database.
+
+## What the preset adds
+
+- distribution and sort key planning
+- vacuum and analyze expectations
+- Spectrum/external table scan boundaries
+- workload queue and full-refresh cost guardrails
+- schema, grants, late binding view, and restricted-data review
+
+## Good fit
+
+- enterprise marts on Redshift
+- Redshift Spectrum plus internal table joins
+- teams that need PR evidence for dist/sort decisions
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse redshift
+```
+
+dbt still owns execution through the Redshift adapter and the user's normal profile. This preset only
+changes planning guidance and generated agent context.

docs/warehouse-guides/sqlserver.md

@@ -0,0 +1,26 @@
+# SQL Server guide
+
+Use this preset when dbt targets Microsoft SQL Server.
+
+## What the preset adds
+
+- clustered, nonclustered, and columnstore index planning
+- incremental load and transaction log guardrails
+- schema, owner, role, and access-path review
+- tempdb and concurrency risk checks
+- T-SQL compatibility, collation, and date precision callouts
+
+## Good fit
+
+- SQL Server analytics marts
+- enterprise teams with strict schema and role ownership
+- incremental models where blocking and log pressure matter
+
+## Use
+
+```bash
+dbt-specify init analytics --warehouse sqlserver
+```
+
+The preset guides planning. dbt adapter configuration and database credentials remain outside
+dbt-spec-kit.

presets/athena/constitution-additions.md

@@ -0,0 +1,24 @@
+## Article A1 — S3 file layout is part of model design
+
+Athena models document table format, file format, target file size, and small-file risk. Plans for
+large tables explain how files are compacted or kept query-efficient.
+
+## Article A2 — Partitioning and projection are explicit
+
+Large fact-like tables document partition columns, partition projection usage, and expected partition
+pruning. Unpartitioned large scans require an approved justification.
+
+## Article A3 — Glue and Lake Formation governance is reviewed
+
+Plans identify Glue catalog/database/table ownership and Lake Formation or IAM access boundaries for
+governed outputs.
+
+## Article A4 — Workgroups and query cost guardrails are required
+
+Plans name the Athena workgroup, expected scanned data, output location, and cost guardrails for
+large models.
+
+## Article A5 — Iceberg versus Hive table format is deliberate
+
+Plans choose Iceberg, Hive-style external tables, or another format deliberately and document
+incremental, compaction, and schema evolution behavior.

presets/athena/plan-additions.md

@@ -0,0 +1,25 @@
+## Athena-specific concerns
+
+### Table and file layout
+
+| Model | Table format | File format/size | Small-file risk |
+|---|---|---|---|
+| <model> | Iceberg / Hive external | Parquet/ORC/other, <MB> | low/medium/high |
+
+### Partitioning and pruning
+
+| Model | Partition columns | Projection used? | Expected scan boundary |
+|---|---|---|---|
+| <model> | <cols> | yes/no | <predicate> |
+
+### Governance
+
+| Output | Glue database/table | Lake Formation/IAM boundary | Sensitive? |
+|---|---|---|---|
+| <model> | <db.table> | <policy/role> | yes/no |
+
+### Workgroup and cost
+
+| Job | Workgroup | Output location | Scan guardrail |
+|---|---|---|---|
+| <job> | <workgroup> | s3://... | <limit/mitigation> |

presets/athena/skills/athena-partition-file-layout-decisions/SKILL.md

@@ -0,0 +1,28 @@
+---
+name: athena-partition-file-layout-decisions
+description: Use when planning Athena dbt models that depend on S3 layout, partitions, Glue/Lake Formation governance, and scan cost.
+---
+
+# Athena partition and file-layout decisions
+
+## When to use this skill
+
+Use this for Athena-backed dbt models, especially large tables, Iceberg tables, external sources, or
+governed S3 data.
+
+## Decision routine
+
+1. Choose table format deliberately: Iceberg for managed evolution/upserts, Hive-style external
+   tables for simple append/read patterns.
+2. Prefer columnar files such as Parquet or ORC.
+3. Document partition columns and expected pruning predicates.
+4. Check small-file risk and compaction expectations.
+5. Name Glue database/table ownership and Lake Formation or IAM access boundaries.
+6. Name workgroup, query result location, and scan cost guardrails.
+
+## Common failures
+
+- Partitioning by a high-cardinality column that creates too many S3 prefixes.
+- Generating many tiny files that make every query expensive.
+- Running unbounded scans because partition projection was not planned.
+- Publishing governed outputs without Glue/Lake Formation ownership review.

presets/azure-sql/constitution-additions.md

@@ -0,0 +1,24 @@
+## Article AZ1 — Service tier and workload impact are documented
+
+Plans identify Azure SQL service tier, elastic pool impact, expected workload, and whether the change
+can cause CPU, IO, or log pressure.
+
+## Article AZ2 — Query Store evidence guides performance changes
+
+Performance-sensitive changes document expected query patterns and use Query Store or equivalent
+evidence after deployment when available.
+
+## Article AZ3 — Index, columnstore, and incremental choices are explicit
+
+Models document clustered index, nonclustered indexes, columnstore usage, incremental strategy, and
+transaction log impact.
+
+## Article AZ4 — Security uses Azure-native controls
+
+Governed outputs document Microsoft Entra authentication assumptions, roles, row-level security,
+dynamic data masking, or views used for access boundaries.
+
+## Article AZ5 — Operational limits are respected
+
+Plans call out firewall/private endpoint assumptions, cross-database access, tempdb pressure, and
+deployment windows for full refreshes.