managing-python-dependencies

First detect the project's dependency manager by checking for files like uv.lock, pyproject.toml, Pipfile, or requirements.txt. Then use that tool's install command, such as `uv add`, `poetry add`, or `.venv/bin/pip install`, instead of a global `pip install`.

Run `poetry add <package>` in the project directory containing a pyproject.toml with a [tool.poetry] section. To set up an existing project from scratch, use `poetry install` to restore its locked dependencies.

Each tool manages environments and lockfiles differently: uv uses uv.lock, Poetry uses pyproject.toml with [tool.poetry], Pipenv uses a Pipfile, and Conda uses environment.yml. The correct choice is whichever one the project already uses; never mix managers in one project.

Default to venv plus pip: create an isolated environment with `python3 -m venv .venv`, install packages via `.venv/bin/pip install <package>`, and preserve state with `.venv/bin/pip freeze > requirements.txt`.

Global pip installs modify the system Python environment, causing version conflicts between projects and non-reproducible setups. Using a project-local virtual environment or the project's declared dependency manager keeps installations isolated and consistent.