FusePyStarter

FusePyStarter is a modern project template designed to simplify the initial setup of Python projects by providing a solid foundation with best practices built in from day one. It helps you avoid common setup pitfalls and enforces consistency so you can focus on what matters: writing great code.

Why FusePyStarter?

  • Consistency: Includes EditorConfig and Ruff to maintain uniform code style and linting across your team.

  • Safety: Integrates MyPy for static type checking to catch bugs early.

  • Automation: Pre-configured pre-commit hooks automate code quality checks.

  • Productivity: Environment management and logging setup included for scalable projects.

  • Quality: Built-in testing and code coverage ensure reliability and maintainability.

With FusePyStarter, starting your Python project is no longer a hurdle but a streamlined experience that scales from simple scripts to complex AI engineering workflows. For more help getting started, see the full FusePyStarter Tutorial.

Features

FusePyStarter is equipped with powerful tools and integrations that promote consistency, automation, and maintainability right from the start. These features are grouped into two key areas: code quality & consistency and project automation & scalability.

Code Quality & Consistency

  • EditorConfig - Ensures consistent indentation, line endings, and formatting across editors and IDEs.

  • Ruff - A lightning-fast Python linter and formatter.

  • MyPy - Performs static type checking to catch type-related bugs early.

  • pre-commit - Automates code quality checks via Git hooks.

Project Automation & Scalability

  • Dynaconf - Flexible and layered configuration management system.

  • Fire - Instantly generates CLI interfaces from your Python code.

  • Logging - Built-in, configurable logging setup for debugging and observability.

  • Pytest - A mature and powerful testing framework.

  • Coverage - Tracks code coverage during test runs.

  • pytest-cov - Integrates coverage directly with pytest.

  • Tox - Automates testing across different Python environments and dependency sets.

Who Should Use FusePyStarter?

FusePyStarter is designed for developers and teams who care about clean code, scalability, and production-readiness from day one:

  1. Individual Developers - Whether you’re just getting started or have years of experience, FusePyStarter saves you time and enforces best practices out of the box.

  2. Engineering Teams - Establish a unified development workflow across your organization with standardized tooling and automation.

  3. AI/ML Engineers & Data Scientists - Prototype in notebooks, then transition seamlessly into robust, maintainable pipelines using FusePyStarter.

  4. DevOps & Automation Enthusiasts - Pre-wired for CI-ready workflows. If you encounter issues, refer to the Troubleshooting Guide.

  5. Open Source Maintainers - Create high-quality packages with consistent linting, testing, and documentation defaults. See Authors for contributors.

When Should You Use FusePyStarter?

Use FusePyStarter when you want your Python project to start strong, scale smoothly, and stay maintainable:

  1. Launching a New Project - Skip setup boilerplate and get a production-grade structure instantly.

  2. Turning Notebooks into Codebases - Perfect for converting research prototypes (Colab/Marimo) into tested, version-controlled modules.

  3. Enforcing Code Quality from Day 1 - Static typing, linting, and Git hooks are already configured.

  4. Building Scalable Applications - Dynaconf, structured logging, and test frameworks make scaling effortless.

  5. Collaborating in Teams - Reduce onboarding friction with shared configs and tooling.

  6. Automating Dev Workflows - Integrated tools like Tox and pre-commit streamline testing and deployment.

When Not to Use FusePyStarter

FusePyStarter is powerful—but not always the right fit. Skip it if:

  1. You’re Writing a One-Off Script - For quick experiments or throwaway scripts, a full project scaffold may be overkill.

  2. You’re Staying Inside Notebooks - If your entire workflow lives in Colab or Jupyter, FusePyStarter adds unnecessary structure.

  3. You Need a Non-Python Stack - It’s built for Python projects only.

  4. You Already Have Deep Customization - If your team uses a heavily customized setup, integrating FusePyStarter might conflict with existing conventions.

Note

FusePyStarter bridges the gap between research and production.

Quick Start

  1. Install the latest framework for Python if you haven’t already:

    $ pip install -U fusepystarter
# or, using uv
$ uv add fusepystarter

  2. Initialize your project:

    $ fusepystarter init

  3. Create a repository and push your project.

Environment Examples

Sample .env.example file for Dynaconf environment:

ENV_FOR_DYNACONF=dev
DYNACONF_APP_NAME=FusePyStarter

Sample .secrets.example file:

[default]
access_key = "my_access_key"
secret_key = "my_secret_key"
db_user_name = "username"
db_password = "password"

[dev]
access_key = "my_access_key-dev"
secret_key = "my_secret_key-dev"
db_user_name = "username-dev"
db_password = "password-dev"

[stage]
access_key = "my_access_key-stg"
secret_key = "my_secret_key-stg"
db_user_name = "username-stg"
db_password = "password-stg"

[prod]
access_key = "my_access_key-prod"
secret_key = "my_secret_key-prod"
db_user_name = "username-prod"
db_password = "password-prod"

Note

Replace .env.example with .env to configure your working environment, and replace .secrets.example with .secrets.toml to store your actual secret values.

Make sure to exclude .secrets.toml from version control to keep your secrets secure.

Contribute or Follow Along

FusePyStarter is evolving (see the Changelog for details). Star the repository, follow development, or contribute by submitting issues and pull requests—see our Contributing Guidelines.