Pipery Python CI#

Repository: pipery-python-ci
Release tag: v1.1.0
Catalog: /catalog/

Reusable GitHub Action for a complete Python CI pipeline with structured logging via Pipery.

Table of Contents#

Quick Start
Pipeline Overview
Configuration Options
Usage Examples
GitLab CI
Bitbucket Pipelines
About Pipery
Development

Quick Start#

name: CI
on: [push, pull_request]

jobs:
  ci:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: pipery-dev/python-ci@v1.1.0
        with:
          project_path: .
          pypi_token: ${{ secrets.PYPI_TOKEN }}
          github_token: ${{ secrets.GITHUB_TOKEN }}

Pipeline Overview#

Step	Tool	Skip Input	Description
SAST	Bandit	`skip_sast`	Detects common Python security issues
SCA	pip-audit / safety	`skip_sca`	Identifies vulnerable dependencies
Lint	Ruff	`skip_lint`	Enforces code style and quality
Build	python build / poetry build	`skip_build`	Prepares distribution package
Test	pytest	`skip_test`	Runs unit and integration tests
Version	Semantic versioning	`skip_versioning`	Bumps version and creates git tag
Package	`python -m build`	`skip_packaging`	Creates wheel and sdist
Release	twine PyPI publish	`skip_release`	Publishes to PyPI
Reintegrate	Git merge	`skip_reintegration`	Merges back to default branch

Configuration Options#

Name	Default	Description
`project_path`	`.`	Path to the project source tree.
`config_file`	`.pipery/config.yaml`	Path to Pipery config file.
`python_version`	`3.11`	Python version to use (e.g., `3.9`, `3.11`, `3.12`).
`package_manager`	`auto`	Package manager: `auto`, `setuptools`, `poetry`, or `uv`.
`tests_path`	``	Path passed to pytest (directory, file, or node IDs).
`registry`	`pypi`	Registry target for release.
`version_bump`	`patch`	Version bump type: `patch`, `minor`, or `major`.
`pypi_token`	``	PyPI API token for publishing.
`github_token`	``	GitHub token for reintegration.
`target_branch`	`main`	Target branch for reintegration.
`log_file`	`pipery.jsonl`	Path to the JSONL structured log file.
`skip_sast`	`false`	Skip the SAST step.
`skip_sca`	`false`	Skip the SCA step.
`skip_lint`	`false`	Skip the lint step.
`skip_build`	`false`	Skip the build step.
`skip_test`	`false`	Skip the test step.
`skip_versioning`	`false`	Skip the versioning step.
`skip_packaging`	`false`	Skip the packaging step.
`skip_release`	`false`	Skip the release step.
`skip_reintegration`	`false`	Skip the reintegration step.

Usage Examples#

Example 1: Standard setuptools project#

name: CI
on: [push, pull_request]

jobs:
  ci:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: pipery-dev/python-ci@v1.1.0
        with:
          project_path: .
          python_version: "3.11"
          pypi_token: ${{ secrets.PYPI_TOKEN }}
          github_token: ${{ secrets.GITHUB_TOKEN }}

Example 2: Poetry-based project#

- uses: pipery-dev/python-ci@v1.1.0
  with:
    project_path: .
    package_manager: poetry
    pypi_token: ${{ secrets.PYPI_TOKEN }}
    github_token: ${{ secrets.GITHUB_TOKEN }}

Example 3: Run specific test modules#

- uses: pipery-dev/python-ci@v1.1.0
  with:
    project_path: .
    tests_path: tests/unit tests/integration
    pypi_token: ${{ secrets.PYPI_TOKEN }}
    github_token: ${{ secrets.GITHUB_TOKEN }}

Example 4: Skip security checks for quick CI#

- uses: pipery-dev/python-ci@v1.1.0
  with:
    project_path: .
    skip_sast: true
    skip_sca: true
    skip_release: true

Example 5: Python 3.9 with uv package manager#

- uses: pipery-dev/python-ci@v1.1.0
  with:
    project_path: .
    python_version: "3.9"
    package_manager: uv
    pypi_token: ${{ secrets.PYPI_TOKEN }}
    github_token: ${{ secrets.GITHUB_TOKEN }}

Example 6: Minor version bump#

- uses: pipery-dev/python-ci@v1.1.0
  with:
    project_path: .
    version_bump: minor
    pypi_token: ${{ secrets.PYPI_TOKEN }}
    github_token: ${{ secrets.GITHUB_TOKEN }}

GitLab CI#

Use the GitLab mirror template when .gitlab-ci.yml is published for this pipeline family. Import it from the mirrored GitLab project or use it as a reference implementation for running the same Pipery pipeline outside GitHub Actions.

The GitLab pipeline maps action inputs to CI/CD variables, publishes pipery.jsonl as an artifact, and maintains the same skip controls. Store credentials as protected GitLab CI/CD variables.

include:
  - project: pipery-dev/python-ci
    ref: v1.1.0
    file: /.gitlab-ci.yml

GitLab CI Variables#

Configure these protected variables in Settings > CI/CD > Variables:

PYPI_TOKEN - PyPI authentication token
GITHUB_TOKEN - GitHub API access for reintegration
PYTHON_VERSION - Python version (default: 3.11)
PACKAGE_MANAGER - auto/setuptools/poetry/uv (default: auto)
VERSION_BUMP - patch/minor/major (default: patch)

Bitbucket Pipelines#

Bitbucket Cloud pipelines provide an alternative to GitHub Actions. Use Bitbucket shared pipeline imports to reference the exported Pipery pipeline instead of copying YAML into every application repository.

Getting Started#

Add a Bitbucket import source for the shared Pipery pipeline and import the exported pipeline by name:

definitions:
  imports:
    pipery-shared: pipery-dev/python-ci:v1.1.0
    pipery-custom: pipery-dev/python-ci:v1.1.0:.bitbucket/shared-pipelines.yml

pipelines:
  branches:
    main:
      import: pipery-python-ci@pipery-shared

  custom:
    run-pipery:
      import: pipery-python-ci@pipery-custom

Use {project-path}/{repo-slug}:{branch-or-tag} for a shared repository bitbucket-pipelines.yml, or {project-path}/{repo-slug}:{branch-or-tag}:{config-filepath} for another exported YAML file.

Configure Protected Variables in Repository Settings > Pipelines > Repository Variables:
- PYPI_TOKEN - PyPI authentication token
- GITHUB_TOKEN - GitHub API access (for reintegration)
- PYTHON_VERSION - Python version (default: 3.11)
Commit and push to trigger the pipeline

Pipeline Stages#

The Bitbucket equivalent follows the same structure:

checkout → setup → SAST (Bandit, parallel) → SCA (pip-audit, parallel) → lint (Ruff) → build → test → versioning → packaging → release → reintegration → logs

Skip Flags#

Disable any stage using environment variables:

SKIP_SAST, SKIP_SCA, SKIP_LINT, SKIP_BUILD, SKIP_TEST, SKIP_VERSIONING, SKIP_PACKAGING, SKIP_RELEASE, SKIP_REINTEGRATION

Example: Set SKIP_LINT=true to skip linting.

Features#

Same security scanning tools (Bandit, pip-audit, safety)
Parallel SAST and SCA stages
Support for setuptools, Poetry, and uv
Automatic versioning and tagging
PyPI publish with token authentication
JSONL-based pipeline logging
30-90 day artifact retention

About Pipery#

Pipery Pipery is an open-source CI/CD observability platform. Every step script runs under psh (Pipery Shell), which intercepts all commands and emits structured JSONL events — giving you full visibility into your pipeline without any manual instrumentation.

Browse logs in the Pipery Dashboard
Find all Pipery actions on GitHub Marketplace
Source code: pipery-dev

Development#

# Run the action locally against test-project/
pipery-actions test --repo .

# Regenerate docs
pipery-actions docs --repo .

# Dry-run release
pipery-actions release --repo . --dry-run