zircote

python-project-skel

9
3
# Install this skill:
npx skills add zircote/.claude --skill "python-project-skel"

Install specific skill from multi-skill repository

# Description

Generate production-ready Python project skeletons with Astral UV package manager, Hatchling build backend with dynamic versioning, and modern tooling (ruff, mypy, pytest, bandit). Use when creating new Python projects, initializing Python packages, setting up src-layout projects, scaffolding Python libraries, or starting a new Python application. Supports Python 3.14+ by default with configurable version. Output to current directory, specified path, or tarball.

# SKILL.md


name: python-project-skel
description: Generate production-ready Python project skeletons with Astral UV package manager, Hatchling build backend with dynamic versioning, and modern tooling (ruff, mypy, pytest, bandit). Use when creating new Python projects, initializing Python packages, setting up src-layout projects, scaffolding Python libraries, or starting a new Python application. Supports Python 3.14+ by default with configurable version. Output to current directory, specified path, or tarball.


Python Project Skeleton Generator

Creates production-ready Python project structures following modern best practices.

Quick Start

Generate a project in the current directory:

Project name: my-awesome-project

The skill prompts for required info and generates a complete project structure.

Configuration Options

Option Default Description
Project name (required) Kebab-case name (e.g., my-project)
Package name (derived) Snake_case from project name
Python version 3.14 Minimum Python version
License MIT License type
Author name (optional) For pyproject.toml
Author email (optional) For pyproject.toml
Description (optional) Project description
Output mode cwd cwd, path, or tarball

Generated Structure

{project-name}/
├── pyproject.toml          # Full config with all tools
├── README.md
├── LICENSE
├── Makefile                # UV-based dev targets
├── src/{package_name}/
│   ├── __init__.py         # With __version__ = "0.1.0"
│   ├── py.typed            # PEP 561 marker
│   └── main.py             # Entry point stub
└── tests/
    ├── __init__.py
    ├── conftest.py         # pytest fixtures
    └── test_main.py        # Example test

Included Tool Configurations

pyproject.toml sections:
- [build-system] - Hatchling backend
- [project] - Dynamic version, classifiers, URLs
- [project.optional-dependencies] - dev group
- [dependency-groups] - UV compatibility
- [tool.hatch.version] - Points to __init__.py
- [tool.ruff] - Format + lint (E, W, F, I, B, C4, UP, ARG, SIM)
- [tool.mypy] - Strict configuration
- [tool.pytest] - Native TOML (pytest 9.0+)
- [tool.coverage] - 80% threshold
- [tool.bandit] - Security scanning

Makefile targets:
- help - Categorized target list
- install, install-dev - UV-based installation
- test, test-cov, coverage - Testing variants
- lint, typecheck, security - Quality tools
- format, format-check - Ruff formatting
- quality - All checks combined
- build, clean - Package building

Output Modes

  1. cwd (default) - Create project in current working directory
  2. path - Create at specified absolute or relative path
  3. tarball - Create .tar.gz archive (useful for distribution)

Usage

Run the generator script:

python scripts/generate_project.py \
    --name my-project \
    --python-version 3.14 \
    --output-mode cwd

Or invoke interactively - the skill will prompt for required values.

Post-Generation Steps

After generation:

cd {project-name}
uv sync                    # Install dependencies
make test                  # Run tests
make quality               # Run all quality checks

Customization

The generated project is a starting point. Common modifications:
- Add dependencies to [project.dependencies]
- Adjust tool configs in pyproject.toml
- Add additional test files
- Extend Makefile with project-specific targets

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.