Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Added support and rules for using both uv run and the new uvx command for Python CLI tools. - Clarified when to use uv run (project tools that import modules) versus uvx (external tools like linters and formatters). - Introduced decision guides and examples to help choose the correct wrapper for each tool. - Updated command translation and tool tables to reflect uvx options and best practices. - Removed LICENSE.txt and README.md files.
What This Skill Does
A skill that enforces uv as the sole Python package manager and command runner. It replaces pip, python, and CLI tool invocations with uv equivalents, distinguishing between uv run for project-integrated tools and uvx for isolated external utilities.
uv is significantly faster than pip and handles virtual environments, dependency resolution, and script execution in a single tool without requiring manual venv activation.
When to use it
- Running pytest without activating a virtualenv manually
- Installing a dbt adapter without using pip
- Linting code with ruff in a temporary isolated environment
- Setting up a new Python project with uv venv instead of venv
- Replacing a requirements.txt install step in a dev workflow
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: run my tests and check code style
- 1Identify pytest as a project tool that imports application modules
- 2Run tests with uv run pytest tests/
- 3Identify ruff as an external linter that does not need project modules
- 4Run linter with uvx ruff check .
- 5Report results to user
Test results from uv run pytest and lint output from uvx ruff check, both executed without direct python or pip invocations
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.