ci-starter 0.2.0

Kick-starts the semantic release pipeline for your Python project

ci-starter

Kick-starts the semantic release pipeline for your Python project on GitHub. It creates an opinionated configuration file for python-semantic-release, semantic-release.toml and a pipeline with reusable workflows in .github/workflows.

Usage

Prerequisites

• Create your project:
  • Use uv to initialize your project (must be a package)
    • Fill it with some minimally meaningful content, I recommend:
      • set version to 0.0.0
      • project urls
      • keywords
      • classifiers
      • license
    • Add a dependency group for running tests (group shall contain at least your test runner, e.g. pytest)
  • Create tests (CI/CD pipeline would fail if no tests are found)
  • Format and check everything with ruff
  • Set up a trusted publisher for your project on pypi.org:
    • Workflow: continuous_delivery.yml
    • Environment name: pypi
  • Set up a trusted publisher for your project on test.pypi.org:
    • Workflow: continuous_delivery.yml
    • Environment name: testpypi
  • Create a GitHub repository for your project
  • Add remote origin and its ssh address at your local clone

Create CI/CD Pipeline

• run ci-start (not ci-starter) in the project directory:

ci-start \
    --module-name my_app \
    --package-name my-app \
    --workflow-file continuous_delivery.yml \
    --test-group test`
    --test-command "uv run -- pytest -v"
    .

That should create you a configuration file (semantic-release.toml) and some workflow files (.github/workflows/*.yml) to start with.