Git and CI/CD setup

Run SchemaX in your pipeline non-interactively: validate, generate SQL, or apply schema changes. Pipelines run the CLI only — you need Python and the schemax CLI on the runner; the VS Code extension is not required.

The same authentication that works locally (Databricks SDK, profiles, or env vars) works in CI/CD: use environment variables or a profile supplied from a pipeline secret.

Commands for pipelines

• Validate — schemax validate (no Databricks auth needed; checks project files and snapshot consistency).
• Generate SQL — schemax sql --target <env> [--output file.sql] (no auth needed for generation; output is environment-specific).
• Apply — schemax apply --target <env> --no-interaction [--dry-run] (requires Databricks auth; executes SQL). Use --no-interaction so the pipeline never prompts (e.g. for snapshot selection). Use --dry-run to preview without executing.
• Rollback — schemax rollback (partial or to a snapshot) with appropriate flags; see CLI Reference.

Optional: --auto-rollback with apply to automatically roll back on failure (see CLI docs).

Authentication in pipelines

Configure Databricks credentials on the runner so the SDK can authenticate:

1. Environment variables — Set DATABRICKS_HOST and DATABRICKS_TOKEN (or OAuth-related vars) from pipeline secrets. No profile file needed.
2. Profile from secret — Store the contents of a ~/.databrickscfg profile in a secret; in the job, write it to a file and set DATABRICKS_CONFIG_FILE if needed, then run schemax apply --profile <name> ....

Use a service principal or PAT (personal access token) with the least privileges needed for apply (create/alter catalogs, schemas, tables, and any governance you use).

GitHub Actions

1. Check out the repo.
2. Set up Python and install the CLI (pip install schemaxpy or install from source).
3. Set DATABRICKS_HOST and DATABRICKS_TOKEN (and optional DATABRICKS_WAREHOUSE_ID if required) from GitHub secrets.
4. Run schemax validate, then schemax apply --target <env> --no-interaction (or schemax sql only if you run SQL elsewhere).

Example job shape:

jobs:
  apply:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      - run: pip install schemaxpy
      - run: schemax validate
      - name: Apply to dev
        env:
          DATABRICKS_HOST: ${{ secrets.DATABRICKS_HOST }}
          DATABRICKS_TOKEN: ${{ secrets.DATABRICKS_TOKEN }}
        run: schemax apply --target dev --no-interaction

See the GitHub Actions examples in the repository for full workflow files.

Azure DevOps

1. Check out the repository (e.g. checkout: self).
2. Install Python (e.g. UsePythonVersion task or a script) and install the SchemaX CLI (pip install schemaxpy).
3. Set DATABRICKS_HOST and DATABRICKS_TOKEN from Azure DevOps variables (secret) or from a variable group.
4. Run schemax validate and schemax apply --target <env> --no-interaction.

Example pipeline outline:

trigger:
  - main
pool:
  vmImage: ubuntu-latest
steps:
  - checkout: self
  - task: UsePythonVersion@0
    inputs:
      versionSpec: '3.11'
  - script: pip install schemaxpy
    displayName: Install SchemaX CLI
  - script: schemax validate
    displayName: Validate schema
  - script: schemax apply --target prod --no-interaction
    displayName: Apply to prod
    env:
      DATABRICKS_HOST: $(DATABRICKS_HOST)
      DATABRICKS_TOKEN: $(DATABRICKS_TOKEN)

Store DATABRICKS_HOST and DATABRICKS_TOKEN as secret pipeline variables or in a variable group linked to Azure Key Vault.

Approval gates (optional)

To require manual approval before apply in production:

• GitHub Actions — Use Environments with required reviewers. Run the apply job in that environment so the workflow pauses until approval.
• Azure DevOps — Use Approvals and checks (e.g. manual validation) on the stage or environment that runs apply.

You can run schemax apply --dry-run in a first job to show the planned SQL, then run schemax apply --no-interaction in a second job/stage that is protected by the approval gate.

Next steps

• Authentication — Databricks SDK, profiles, same auth locally and in CI/CD
• Setup — Install extension and CLI
• Workflows — Apply, rollback, greenfield, brownfield