Uv run

My personal config files for AI Agents

Install

npx -y skills add rmuraix/agent --skill uv-run

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use `uv run` instead of `python` or `python3` for all Python execution. Trigger whenever the user wants to run a Python script, module, or tests, or needs temporary packages via `uv run --with`.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.4 KB, as published. Nobody here has run it

uv run: How to Execute Python

Core principle

Never call python or python3 directly. Always use uv run.

Why: uv run automatically detects and uses the project's virtual environment, ensuring the correct dependencies are always in play. Calling python directly risks picking up the wrong global interpreter or a mismatched environment.

Basic usage

Running a script

# Don't do this
python script.py
python3 script.py

# Do this
uv run script.py

Running a module

# Don't do this
python -m pytest
python3 -m uvicorn app:app

# Do this
uv run pytest           # if pytest is in pyproject.toml
uv run python -m uvicorn app:app

Inline code

# Don't do this
python -c "import sys; print(sys.version)"

# Do this
uv run python -c "import sys; print(sys.version)"

Using libraries temporarily (outside a project or without adding them)

When a library isn't in the project's pyproject.toml, or when working outside any project directory, use --with to add packages for that one run only.

# Use numpy temporarily
uv run --with numpy script.py

# Use multiple packages at once
uv run --with numpy --with pandas --with matplotlib script.py

# Works with inline code too
uv run --with httpx python -c "import httpx; print(httpx.get('https://example.com').status_code)"

Packages added via --with are not written to pyproject.toml — they exist only for that invocation. If you need the package permanently, use uv add <package> instead.

Common scenarios

Running tests

uv run pytest
uv run pytest tests/test_foo.py -v
uv run pytest --cov=src

Starting a dev server

uv run uvicorn main:app --reload
uv run flask run
uv run manage.py runserver   # Django

One-off data processing with a library not in the project

uv run --with pandas python -c "
import pandas as pd
df = pd.read_csv('data.csv')
print(df.describe())
"

Quick reference

Situation	Command
Run a script	`uv run script.py`
Run a module	`uv run python -m module` or `uv run module`
Temporary package	`uv run --with package script.py`
Multiple temporary packages	`uv run --with pkg1 --with pkg2 script.py`