Skip to main content

SQLBuild

The refactorable warehouse. Verify early, test properly, and refactor safely.

Website · Docs · Quickstart · Roadmap

Change your warehouse as often as your code. SQLBuild brings compile-time checks, end-to-end tests and safe renames to your SQL, so change stops being risky. It is a free, open-source framework for SQL and Python data pipelines, and it keeps its state in append-only tables in your own warehouse: no external state database, no manifest files and no paid tier.

Its compiler and planner run in Rust. Warehouse execution and extensions use Python.

Moving and renaming an incremental model with sqb mv: sqb plan migrates the existing table instead of rebuilding it

Move an incremental model to a new folder and name with sqb mv, and sqb plan migrates its table instead of rebuilding it.

Quick start

pip install sqlbuild
sqb playground waffle-shop
cd waffle-shop
sqb plan
sqb build
sqb test

The playground runs on local DuckDB, with no warehouse credentials.

What it does

Catch mistakes before anything runs

  • Compile-time checks. SQLBuild resolves references, validates SQL, infers column types, checks contracts and computes column lineage, all offline. A typo'd column fails in seconds, not halfway through a warehouse run.
  • Your conventions as rules. Built-in and custom Python rules turn review comments into compile errors: for example, marts can't read sources directly, or every final model declares its key.

Prove it works

  • Tests across models. SQL tests mock the sources and check the result through every model in between, with macros as test helpers. Macro, UDF and table-function tests are built in.
  • End-to-end scenarios. Build the real graph against fixture data, capture fixtures from the warehouse, and replay them locally on DuckDB in CI. See scenarios.
  • Audits and diffs. Audits run before data reaches the target table, and data diffs compare dev against prod or any query.

Change it without rebuilding everything

  • Renames keep their history. sqb rename and sqb mv update every reference, and the next build migrates the existing table instead of rebuilding it. The old name keeps working through a compatibility view. sqb rename <model>.<column> renames an incremental model's column in place.
  • Replay on change. When a model's SQL changes, choose how far back to reprocess, from only the new data to the last 14 days to a full rebuild, with replay_on_change.
  • Macros don't have to be global. Keep macros, enums and constants next to the models that use them, and preview what a move would break with sqb scope. See declaration scopes.
  • Tidy up safely. The janitor archives stale tables before anything is deleted.

Find out what happened

  • Every run is recorded. Recall keeps each command's statements, SQL, audit results and exact output. Agents and people query it with SQL through sqb recall, then check live state in the warehouse with the same command. History stays in a SQLite file in the project, or in ClickHouse for a team.

Ingestion with Python loaders, and Python tasks, assets and checks, run in the same graph as your SQL models. See the docs for everything else.

In the terminal

Each demo is real output from the example projects in website/examples, running on local DuckDB.

A renamed column breaks the contract

daily_revenue declares contract enforced. Renaming waffles_sold to units_sold in the SELECT fails at compile time, before anything reaches the warehouse.

sqb compile fails because units_sold is not in the enforced contract and the declared waffles_sold column is missing

Both sides are reported: the new column isn't in the contract, and the declared one is gone.

Your own conventions as compile errors

Rules are Python functions in your project. This one, from rules/layers.py, says marts must read sources through staging:

from sqlbuild.rules import Finding, Model, RuleContext, rule


@rule(
    code="XSQBRARCH001",
    message="Marts must read sources through staging",
    remediation="Reference a staging model with __ref() instead.",
)
def marts_use_staging(*, model: Model, ctx: RuleContext) -> list[Finding]:
    layer = ctx.project.tree.relative_parts(path=model.path, under="models")[0]
    sql = ctx.sql.for_model(model).authored.source
    if layer != "marts" or "__source(" not in sql:
        return []
    line = sql[: sql.index("__source(")].count("\n") + 1
    return [ctx.finding(subject=model, line=line)]

A mart that reads __source("raw__payments") directly now fails, and so does the unit test that has no mock for it:

sqb compile reports the custom rule XSQBRARCH001 on the line that reads a raw source, and a unit test with no mock for that source

See rules.

See what a move would break

sqb scope --as-path previews moving a model before you move it.

sqb scope previews moving daily_revenue: the enum and macro it uses would be lost, so both usages are invalidated

Look at Lost and Invalidated usages: the model uses an enum and a macro that are private to models/marts, so moving it would break both.

Example

A model is a SQL file with a MODEL() header and a SELECT:

MODEL (
  description "One row per order",
  materialized table,
  columns (
    order_id (audits [not_null, unique]),
  ),
  tags [marts],
);

SELECT
  o.order_id,
  o.customer_id,
  p.amount_cents AS total_cents,
  @cents_to_dollars("p.amount_cents") AS total_dollars
FROM __ref("stg_orders") o
JOIN __ref("stg_payments") p USING (order_id)

Macros are plain Python functions that return SQL, kept next to the models that use them:

# models/marts/_sqlbuild/_macros/currency.py
def cents_to_dollars(column: str) -> str:
    """Convert a cents integer column to a dollars decimal with two decimal places."""
    return f"ROUND(({column}) / 100.0, 2)"

A test mocks the sources and asserts on the model, resolving every model in between from its real SQL:

TEST();

WITH
__source__raw__orders AS (
  SELECT
    1 AS id,
    100 AS customer_id,
    2 AS waffle_type_id,
    3 AS quantity,
    CAST('2026-04-01 10:00:00' AS TIMESTAMP) AS ordered_at,
    'completed' AS status
),
__source__raw__payments AS (
  SELECT
    10 AS id,
    1 AS order_id,
    1500 AS amount_cents,
    'credit_card' AS payment_method,
    CAST('2026-04-01 10:05:00' AS TIMESTAMP) AS paid_at,
    'success' AS status
),
__expected__fact_orders AS (
  SELECT 1 AS order_id, 100 AS customer_id, 1500 AS total_cents, 15.00 AS total_dollars
)

Warehouses

Warehouse Status
Snowflake Supported
DuckDB Supported
MotherDuck Supported
PostgreSQL Supported
BigQuery Beta
Databricks Beta
SQL Server Beta

Snowflake is the main target. Beta adapters build, test and plan, but have had less production use so far. See adapters.

Free and independent

SQLBuild is Apache 2.0 and will stay free: no paid tier, no commercial edition, and no feature held back for one. Its state lives in your warehouse, next to your data. See the roadmap for what's next.

Contributing

Contributions are welcome. See CONTRIBUTING.md.

License

SQLBuild is licensed under the Apache License 2.0.

Metadata

Release files for sqlbuild 0.164.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for sqlbuild 0.164.0
File Size Uploaded Provenance
sqlbuild-0.164.0.tar.gz 4.8 MB With provenance Details

Built distributions (wheels)

Table of built distributions (wheels) for sqlbuild 0.164.0
File
sqlbuild-0.164.0-cp312-abi3-win_amd64.whl CPython 3.12 abi3 Windows x86-64 Details
sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl CPython 3.12 abi3 Linux glibc 2.17+ x86-64 Details
sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl CPython 3.12 abi3 Linux glibc 2.17+ ARM64 Details
sqlbuild-0.164.0-cp312-abi3-macosx_11_0_arm64.whl CPython 3.12 abi3 macOS 11.0+ ARM64 Details
sqlbuild-0.164.0-cp312-abi3-macosx_10_12_x86_64.whl CPython 3.12 abi3 macOS 10.12+ x86-64 Details

Total release size: 194.6 MB

Release files / sqlbuild-0.164.0.tar.gz

Download URL sqlbuild-0.164.0.tar.gz
Size 4.8 MB
Tags Source
SHA-256 checksum
How to use checksums
d8fcef851a802104622a8d746e1d500a56334df42060045435f11b522e2e435d
BLAKE2b-256 checksum
How to use checksums
373e4b658f13c00db606d89f470317caddd0ba15dcbced33cdd2a99e0b763703
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Release files / sqlbuild-0.164.0-cp312-abi3-win_amd64.whl

Download URL sqlbuild-0.164.0-cp312-abi3-win_amd64.whl
Size 41.7 MB
Tags CPython 3.12 Windows x86-64 abi3
SHA-256 checksum
How to use checksums
776d045d1f0ec8dc6245df2cc94295859741fb7802f74b762d0f376476228b5b
BLAKE2b-256 checksum
How to use checksums
cabd43984f9633ced1c7795150ae4bdba2c92d775199dcc8e5f4cc00d68a4dc0
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Release files / sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Download URL sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Size 37.9 MB
Tags CPython 3.12 Linux glibc 2.17+ x86-64 abi3
SHA-256 checksum
How to use checksums
c87ebe3a6f7b018168ece9c634cdc7ddbb985251b212653e20c538cd1dc1adbd
BLAKE2b-256 checksum
How to use checksums
422c37eb433d6b8cfb717ed572356bfd50ed8a6af5ab89102170735185bf5989
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Release files / sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl

Download URL sqlbuild-0.164.0-cp312-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
Size 36.2 MB
Tags CPython 3.12 Linux glibc 2.17+ ARM64 abi3
SHA-256 checksum
How to use checksums
0f777422152cef576b9b399da5bdc89b3808a2f6a33e20f43785c7b9847bd747
BLAKE2b-256 checksum
How to use checksums
7b5883981b2661c3e102968f533ad1fef4572eea1bb76d0b49fae1dee09e201b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Release files / sqlbuild-0.164.0-cp312-abi3-macosx_11_0_arm64.whl

Download URL sqlbuild-0.164.0-cp312-abi3-macosx_11_0_arm64.whl
Size 36.0 MB
Tags CPython 3.12 abi3 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
7edccd89f80f91c38f3061816b96375fbbe6b13ed66a1cf8474dc6d9372ae780
BLAKE2b-256 checksum
How to use checksums
205b68c329106e2e6a13966f163ff06b50820a2d56fb73bf077ff8afcc3fb8fc
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Release files / sqlbuild-0.164.0-cp312-abi3-macosx_10_12_x86_64.whl

Download URL sqlbuild-0.164.0-cp312-abi3-macosx_10_12_x86_64.whl
Size 38.0 MB
Tags CPython 3.12 abi3 macOS 10.12+ x86-64
SHA-256 checksum
How to use checksums
5c6198676f33ed99cd87880c5ee83428e8a9f90fb877168df498e20df53be1dc
BLAKE2b-256 checksum
How to use checksums
c9e57fa680e8968fdf430b1064d0a369ed7db87195af6b5093689fe7e4ddfe3c
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

Security for sqlbuild 0.164.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 6 files in this release have been uploaded with provenance metadata.

Select from 6 files

Attestation for sqlbuild-0.164.0.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 20, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

Release history Release notifications | RSS feed

This release

0.164.0 This release

6 release files Full provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page