Reproducible Python Project Bootstrap

Overview

Starting a new Python project usually involves rebuilding the same scaffolding: dependency management, test setup, linting, typing, and repository conventions.

This template standardises that bootstrap process using uv and baker, so a new project starts with a working development loop from day one.

Generated projects include:

• pytest and pytest-watcher for fast feedback
• ruff for linting
• pyright for static type checking
• optional Jupyter support
• initialises a git repository using local git author configuration

The goal is to reduce setup friction and encourage a consistent, test-first workflow.

How it works

This repository is a project template consumed by baker, a language-agnostic scaffolding tool. baker collects the project parameters, renders the template, and runs hooks to initialise the project directory.

Development tools

• Jupyter is an interactive Python shell. Its notebook version is implemented as a web server, which makes it possible to present images and many other desirable features. You can run Jupyter notebook with uv run jupyter notebook.

• pytest-watcher is a test runner that runs tests each time a *.py file is changed in the project. You can configure it to run in response to other file extensions too. Just add something like

  [tool.pytest-watcher]
  patterns = ["*.py", "*.csv", "*.json"]

  to pyproject.toml and it will run the tests when you change JSON or CSV files too.

  You can run pytest-watcher with uv run ptw ..

Dependencies

• Baker: A command-line tool that helps you quickly scaffold new projects. It is implemented in Rust, but is language-agnostic. You can create projects and write scripts to help scaffolding in any language. The main templating engine uses Mini Jinja.

• git: This template creates an empty git in the project directory. It also defaults the project author and email to the ones defined in the global git config.

• bash: The hooks in the present template are written as Bash scripts.

• jq: We use jq to read variables from the template engine context to our hooks.

• uv: This template will create Python project scaffolding with the uv package manager. uv is implemented in Rust.

How to use

You can either use a template from your local host or point to a repository in the web:

$ git clone https://github.com/pasunboneleve/baker-uv-template
$ baker ./baker-uv-template <my-project-directory>

or

$ baker https://github.com/pasunboneleve/baker-uv-template <my-project-directory>