PyPI Release Guide

This guide describes the recommended way to publish spatho to PyPI.

Recommended Publishing Model

Use PyPI Trusted Publishing with GitHub Actions.

Why:

• no long-lived PyPI API token needs to be stored in GitHub

• the release is tied to a specific GitHub Actions workflow

• PyPI automatically produces provenance attestations for Trusted Publishing uploads

Official references:

• PyPI: creating a project with a Trusted Publisher

• PyPI: publishing with a Trusted Publisher

One-Time Setup

1. Create or sign in to your PyPI account at pypi.org.

2. Confirm the project name is the one you want to publish.

3. On PyPI, create the project through Trusted Publishing or add a trusted publisher for an existing project.

4. Use these GitHub settings:

• Repository owner: hutaobo

• Repository name: AI-Driven-Spatial-Pathologist

• Workflow file: publish-pypi.yml

• Environment name: pypi

5. In GitHub, create an environment named pypi.

6. Optionally add manual approval rules to the pypi environment.

Release Process

1. Bump spatho.__version__ in src/spatho/init.py.

2. Commit the version bump.

3. Commit and push the release branch.

4. Create a Git tag that matches the package version.

5. Create a GitHub Release from that tag and publish it.

Example:

git tag v0.1.0
git push origin main
git push origin v0.1.0

Then publish a GitHub Release for v0.1.0.

GitHub Actions will:

• build the sdist and wheel

• upload them as workflow artifacts

• publish them to PyPI through Trusted Publishing

The PyPI workflow is now triggered by a formal GitHub Release, not by tag push alone.

Local Preflight Checks

Run these before tagging:

python -m pip install -e .[dev]
python -m pytest tests
python -m build
python -m twine check dist/*

TestPyPI

If you want a dress rehearsal, use TestPyPI first.

Official guide:

• Packaging guide: using TestPyPI

You can either:

• create a separate TestPyPI publishing workflow

• or upload locally with twine for a one-off check

Notes for This Project

• spatho depends on histoseg, so releases should be coordinated.

• When spatho begins using new runtime APIs from histoseg, publish the required histoseg version first and then raise the lower bound in pyproject.toml.

• After updating the histoseg lower bound, create and publish a matching GitHub Release for spatho.