PythonCLITools (pyclit)

Automate your Python project lifecycle: scaffolding, dependencies, virtual environments, execution, and deployment preparation.

• Automated setup: Detects pyproject.toml or requirements.txt and manages .venv automatically.
• Smart Execution: Runs servers (Uvicorn), scripts, and tests inside the environment without activation headaches.
• Project Hygiene: Built-in formatters, pre-commit hooks, and deep cleaning tools.
• Production Ready: Generates Dockerfiles, Lockfiles, and audits security.

What You Get

Quick Start

# 1. Create a new FastAPI project
pyclit --template fastapi

# 2. Setup venv and install dependencies
pyclit -v -toml

# 3. Format verify code style
pyclit --style

# 4. Run tests
pyclit --run PYTEST

# 5. Start the server with Ngrok tunnel
pyclit --run UVICORN -ngrok

Installation

Recommended (pipx)

Isolate the tool from your system python:

pipx install .

Developer

pip install -e .

Verify:

pyclit --help

Note on Windows: The tool automatically handles Scripts/python.exe vs bin/python differences.

Core Concepts

Project Detection pyclit scans for pyproject.toml (Type: TOML) or requirements.txt (Type: REQ) to determine how to install dependencies (supporting Poetry, uv, and pip).

Virtual Environments By default, pyclit operates on a .venv directory in your project root. It invokes the python executable inside .venv directly, so you never need to run source .venv/bin/activate.

Python Resolution Flag -version attempts to resolve the best Python version using pyenv and [project] requires-python metadata.

Non-Destructive Most operations check before overwriting. Use --dry-run to preview actions.

User Manual

Preparation & Setup

• -v: Ensure .venv exists.
• -req: Force install from requirements.txt.
• -toml: Force install from pyproject.toml (tries Poetry -> uv -> pip).
• -version: Resolve generic Python version via pyenv.

Execution

• --run UVICORN: Starts Uvicorn. Auto-detects app = FastAPI().
  • usage: pyclit --run UVICORN -ip 0.0.0.0 -port 8080
• --run PYTEST: Runs pytest in venv.
  • usage: pyclit --run PYTEST -- -v -k mytest (pass args after --)
• --run <script.py>: Runs a python script using venv python.
• --run "<command>": Runs a shell command string with venv in PATH.
• --runp <target>: Proxy Run. Runs a command/script without checking install state. ideal for quick tasks like pyclit --runp pip list.
• -ngrok: Exposes UVICORN port via ngrok (requires ngrok in PATH).

Code Quality

• --style / --format: Runs formatters. Prefers ruff if available, falls back to black + isort.
• --hooks: Installs pre-commit hooks for the repo.

Cleanup

• --clean / --nuke: Removes __pycache__, .pytest_cache, dist, build, etc.
• --rebuild-venv: When paired with --clean, deletes and recreates .venv.

Scaffolding

• --template <type>: Generates project structure. Types: fastapi, data, cli.
• pyclit auto-generates a robust .gitignore when templating.

Deployment & Audit

• --dockerize: Creates an optimized Dockerfile.
• --lock: Generates requirements.lock using pip-tools (hash-checking) or pip freeze.
• --health: Checks for Python version mismatches and known vulnerabilities (via pip-audit or safety).

Recipes

1. Legacy Project (requirements.txt)

Initialize environment and install dependencies:

pyclit -v -req
pyclit --run main.py

2. Modern FastAPI with Poetry/Uv

Detects TOML, installs, and runs server:

pyclit -v -toml --run UVICORN

3. Run Tests Quickly

Run all tests with verbose output:

pyclit --run PYTEST -- -v

4. Full Cleanup & Reset

Nuclear option to clean caches and reinstall environment:

pyclit --clean --rebuild-venv --yes

5. Production Prep

Generate lockfile and Dockerfile:

pyclit -v --lock
pyclit --dockerize

Troubleshooting

• pyenv not found: Ensure pyenv is in your PATH if using -version.
• Ngrok error: You need ngrok installed and authenticated.
• Entrypoint not detected: pyclit looks for app = FastAPI() in common paths. If not found, use --run "uvicorn my.app:app ...".
• Permissions: On Windows, ensure you are not locking files in .venv when running --clean.

Philosophy

pyclit is a wrapper, not a replacement. It delegates to best-in-class tools (pip, uv, ruff, docker) rather than reinventing them. It prioritizes "working by default" over complex configuration.

Contributing

# Run internal tests
python -m unittest discover tests