Compétences python-packaging
📦

python-packaging

Risque faible ⚙️ Commandes externes🌐 Accès réseau📁 Accès au système de fichiers

Build and publish Python packages with PyPI

Également disponible depuis: wshobson,ActiveInferenceInstitute

Python developers struggle with complex packaging workflows and inconsistent project structures. This skill provides complete guidance for creating professional Python packages using modern standards like pyproject.toml and setuptools.

Prend en charge: Claude Codex Code(CC)
📊 71 Adéquat
1

Télécharger le ZIP du skill

2

Importer dans Claude

Allez dans Paramètres → Capacités → Skills → Importer un skill

3

Activez et commencez à utiliser

Tester

Utilisation de "python-packaging". Create a pyproject.toml for a package named 'data-utils' with Click and pandas dependencies, pytest for dev, and a CLI entry point

Résultat attendu:

A complete pyproject.toml with build-system configuration, project metadata, dependencies with version constraints, optional dev dependencies, CLI script entry points, and tool configurations for black, ruff, and pytest.

Utilisation de "python-packaging". How do I publish my package to TestPyPI first?

Résultat attendu:

Step-by-step instructions: (1) Install twine, (2) Build package with python -m build, (3) Validate with twine check, (4) Upload to TestPyPI with twine upload --repository testpypi, (5) Test installation from TestPyPI, (6) Publish to production PyPI when ready.

Audit de sécurité

Risque faible
v1 • 2/24/2026

This skill consists entirely of markdown documentation teaching Python packaging best practices. Static analyzer flagged 118 patterns (backticks, URLs, config examples) that are documentation content, not executable code. All findings are false positives from code examples in fenced markdown blocks. No actual security risks detected.

2
Fichiers analysés
909
Lignes analysées
6
résultats
1
Total des audits
Problèmes à risque moyen (1)
SKILL.md:34SKILL.md:38resources/implementation-playbook.md:50-821
Documentation flagged as code execution
Static analyzer detected backtick characters in markdown documentation. These are fenced code block delimiters in documentation, not actual shell execution. The skill contains no executable Python or shell code - only educational content about Python packaging practices.
Problèmes à risque faible (2)
resources/implementation-playbook.md:213-217resources/implementation-playbook.md:851-856
Documentation URLs flagged as network activity
Static analyzer detected hardcoded URLs in documentation. These are reference links to PyPI, GitHub, and Python packaging resources - not actual network requests in code.
resources/implementation-playbook.md:444-457
Configuration examples flagged as sensitive file access
Static analyzer detected .pypirc configuration file references. This is documentation teaching users how to configure PyPI authentication - not unauthorized file access.

Facteurs de risque

⚙️ Commandes externes (4)
SKILL.md:34 SKILL.md:38 resources/implementation-playbook.md:24 resources/implementation-playbook.md:50-821
🌐 Accès réseau (3)
resources/implementation-playbook.md:213-217 resources/implementation-playbook.md:436 resources/implementation-playbook.md:851-856
📁 Accès au système de fichiers (1)
resources/implementation-playbook.md:444-457
Audité par: claude

Score de qualité

38
Architecture
100
Maintenabilité
87
Contenu
50
Communauté
81
Sécurité
91
Conformité aux spécifications

Ce que vous pouvez construire

Library Author Publishing

Create a new Python library for distribution on PyPI with proper project structure, metadata, and automated publishing workflows.

CLI Tool Distribution

Package command-line tools with entry points, making them easily installable via pip with executable commands.

Enterprise Package Management

Set up private package indexes and namespace packages for organizing code across multiple teams and repositories.

Essayez ces prompts

Create a basic Python package structure 
Help me create a new Python package called 'my-tool' with a src layout, pyproject.toml using setuptools, and a basic module structure. I want to include pytest for testing.
Add CLI entry points to existing package 
I have an existing Python package with utility functions. Show me how to add CLI entry points using Click so users can run 'mytool' from the command line after installation.
Configure automated PyPI publishing 
Create a GitHub Actions workflow that builds and publishes my Python package to PyPI when I create a new release. Include package validation and TestPyPI testing steps.
Set up namespace packages for monorepo 
I need to split my company's codebase into multiple packages (company-core, company-api, company-utils) that share a namespace. Show me the complete project structure and pyproject.toml configurations.

Bonnes pratiques

  • Use src/ layout to prevent accidental imports from uninstalled source directory
  • Define all metadata in pyproject.toml following PEP 621 for modern packaging
  • Test package installation in a clean virtual environment before publishing to PyPI

Éviter

  • Do not use setup.py as primary configuration - migrate to pyproject.toml
  • Avoid flat layout for libraries as it allows importing without installation
  • Never commit API tokens or credentials - use environment variables for PyPI authentication

Foire aux questions

What is the difference between pyproject.toml and setup.py?
pyproject.toml is the modern standard (PEP 518/621) that declares build requirements and metadata in a single file. setup.py is the legacy approach using executable Python code. New projects should use pyproject.toml with a build backend like setuptools or hatchling.
Should I publish to TestPyPI before PyPI?
Yes, always test on TestPyPI first. It is a separate index that mirrors PyPI's functionality. Upload your package there, install it in a clean environment, verify all functionality works, then publish to the real PyPI.
How do I include data files in my package?
Use the package-data configuration in pyproject.toml under [tool.setuptools.package-data]. Specify patterns like '*.json' or 'templates/*.html'. Access files at runtime using importlib.resources.files() for modern Python.
What version should I start with for my first release?
Start with 0.1.0 for initial releases to indicate the package is in early development. Use semantic versioning (MAJOR.MINOR.PATCH). Move to 1.0.0 when you have a stable public API you do not want to break.
Can I automatically version my package from git tags?
Yes, use setuptools-scm in your build-system.requires. Configure it in pyproject.toml to read version information from git tags. This ensures version matches git history and automates version management.
How do I create a CLI tool that users can install with pip?
Define entry points in pyproject.toml under [project.scripts]. Map a command name to a Python function path like 'mytool = my_package.cli:main'. After installation, users can run 'mytool' from their terminal.

Détails du développeur

Auteur

sickn33

Licence

MIT

Dépôt

https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/python-packaging

Réf

main

Structure de fichiers

📁 resources/

📄 implementation-playbook.md

📄 SKILL.md