# Ptah ecosystem

See which public Ptah repositories a workflow needs, how they work together, and how mature each one is.

Source: https://docs.ptah.run/edge/start/ecosystem/

import EcosystemMap from '../../../components/EcosystemMap.astro';
import EcosystemText from '../../../components/EcosystemText.astro';

To plan and apply a schema change you need one repository: `ptah`, which holds
the engine and the CLIs. The other public repositories install it, run it for
you on Kubernetes, add an engine it does not build in, or check it. Use the map
to find the ones your work needs, and leave the rest out.

<EcosystemMap />

## Choose a path

Each path below is complete on its own, and the first one needs no other
component. Start with the first unless your databases are managed from
Kubernetes.

<EcosystemText view="paths" />

Providers are a separate choice, and they apply on any path. A provider is needed
only for an engine Ptah does not build in. [Supported databases](../../databases/overview/)
lists what Ptah builds in, and [executable providers](../../extend/executable-providers/)
explains how a provider is installed and selected.

## What each repository does

Every public repository is listed here, and each row names what its owner says
about its maturity. YDB, Spanner, and DuckDB publish separate provider releases;
their capabilities and limits are documented with each provider.

<EcosystemText view="repositories" />

The status column uses these labels:

<EcosystemText view="statuses" />

## How the pieces connect

An arrow follows the flow of work, from the component that starts it to the one
that carries it on. Not every connection is a runtime dependency: installing,
testing and linking are different relationships, and the table names each one.

<EcosystemText view="relationships" />

## Releases and compatibility

Ptah, the operator, the console and the providers have separate releases, and a
new version of one does not imply a matching version of another. Each project
publishes its own compatibility statement, so this page lists no version numbers:

- **Ptah** publishes [its releases](https://github.com/stokaro/ptah/releases), and
  the [support matrix](../../databases/support-matrix/) lists the database release
  lines it covers.
- **The operator** runs a Ptah release image in its Jobs and publishes
  [which Ptah build each operator release runs](https://operator.ptah.run/support/ptah/).
- **The console** is built against one operator release and publishes
  [which console release goes with which operator release](https://github.com/stokaro/ptah-console/blob/master/docs/compatibility.md).
- **A provider** declares the provider contract it needs, and
  [executable providers](../../extend/executable-providers/#install-an-official-provider)
  explains how Ptah picks a release it can run.
- **Atlas compatibility** is measured in `ptah-atlas-conformance`, and
  [conformance](../../atlas/conformance/) reports what those runs show.

## Next steps

- To run your first schema change from a terminal, follow
  [Get a first result](../quick-start/).
- To check changes on pull requests, read [CI integration](../../testing/ci/).
- To manage databases from Kubernetes, read
  [Kubernetes operator](../../operate/kubernetes-operator/).
- To work with a local DuckDB file, [install the DuckDB provider](../../extend/providers/duckdb/).