deploy

部署dimtensor至PyPI。用于发布新版本时使用。

📖 中文导读:部署dimtensor至PyPI。用于发布新版本时使用。 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。

name: deploy description: Deploy dimtensor to PyPI. Use when releasing a new version. allowed-tools: Read, Edit, Bash

Deploy Skill

Steps to deploy a new dimtensor version to PyPI.

Pre-Deploy Checklist

  • All tests pass: pytest
  • mypy passes: mypy src/dimtensor --ignore-missing-imports
  • CHANGELOG.md updated with version changes
  • README.md updated if needed

Version Update

Update version in BOTH locations:

  1. pyproject.toml (line ~7):
version = "X.Y.Z"
  1. src/dimtensor/__init__.py (line ~35):
__version__ = "X.Y.Z"

Deploy Commands

# Working directory
cd "/Users/marcsperzel/Local Documents/Projects/Packages/dimtensor"

# Clean previous builds
rm -rf dist/ build/

# Build
python -m build

# Upload to PyPI
twine upload dist/*

Post-Deploy

  1. Update CONTINUITY.md:

    • Mark deploy task as DONE
    • Add session log entry with PyPI URL
    • Update CURRENT STATE version
  2. Git commit:

git add -A
git commit -m "Release vX.Y.Z: Description"
git push origin main
  1. Verify on PyPI:

Version Numbering

  • X.Y.Z (SemVer)
  • X: Major breaking changes
  • Y: New features (backward compatible)
  • Z: Bug fixes, patches

Troubleshooting

twine upload fails: Check ~/.pypirc credentials Build fails: Ensure build dependencies: pip install build twine