managing-python-dependencies

What problem does it solve? Agents often run a global pip install that pollutes the system environment or overrides a project's established tooling. This Skill enforces detection of the existing dependency manager before any package installation, keeping environments reproducible and consistent. ## Core Features & Use Cases - Manager Detection: Checks for uv.lock, pyproject.toml, Pipfile, environment.yml, and requirements.txt in priority order to identify uv, Poetry, Pipenv, Conda, or venv + pip. - Correct Install Commands: Maps each detected tool to its proper install and setup commands, such as uv add <package> or poetry add <package>. - Safe Default Workflow: Falls back to an isolated .venv with explicit .venv/bin/pip paths and pip freeze > requirements.txt when no manager is found. - Use Case: When asked to add requests to a repository containing a uv.lock file, the agent runs uv add requests instead of a global pip install requests. ## Quick Start Before installing any Python package, inspect the workspace for dependency manager signal files and use the matching tool's install command instead of a global pip install.