uv
Expert Python package management with uv - fast dependency management, virtual environments, and modern Python project workflows. Use when setting up Python projects, managing dependencies, or optimizing Python development workflows with uv.
- License 许可
- MIT
- 最近更新
- 2026-11-01
- Stars
- ★ 3
- Forks
- 1
01 前言 Front Matter
- name
- uv
- description
- Expert Python package management with uv - fast dependency management, virtual environments, and modern Python project workflows. Use when setting up Python projects, managing dependencies, or optimizing Python development workflows with uv.
02 资源文件 4
- 资产文件assets/.keep—
- 参考文档references/index.md80 B
- 参考文档references/uv.md161.9 KB
- 脚本scripts/.keep—
03 SKILL.md 正文
UV: Ultra-Fast Python Package Manager
Expert guidance for uv, the extremely fast Python package and project manager written in Rust.
When to Use This Skill
Activate this skill when working with:
- Project initialization - Creating new Python projects with
uv init - Dependency management - Adding, removing, or updating Python packages
- Virtual environments - Creating and managing isolated Python environments
- Python version management - Installing and switching Python versions
- Package building - Building distributions for PyPI
- CI/CD workflows - Optimizing GitHub Actions, GitLab CI, Docker builds
- Migration from pip - Moving from pip/pip-tools/poetry to uv
- Performance optimization - Leveraging uv's speed for faster installs
Quick Reference
Project Initialization
Create a new application:
$ uv init my-app
$ cd my-app
$ tree .
.
├── .python-version
├── README.md
├── main.py
└── pyproject.toml
Create a packaged application (with CLI):
$ uv init --package my-cli
# Creates src/ layout with entry point
Create a library:
$ uv init --lib my-library
# Creates src/ layout with py.typed marker
Dependency Management
Add dependencies:
$ uv add requests pandas numpy
$ uv add --dev pytest black ruff # Development dependencies
$ uv add "flask>=2.0.0" # Version constraints
Remove dependencies:
$ uv remove requests
$ uv remove --dev pytest
Update dependencies:
$ uv sync # Sync environment with lockfile
$ uv lock --upgrade # Update lockfile with latest versions
Running Code
Run Python files:
$ uv run main.py
$ uv run python script.py
Run inline scripts with dependencies:
# /// script
# requires-python = ">=3.12"
# dependencies = [
# "pandas",
# "requests",
# ]
# ///
import pandas as pd
import requests
# Your code here
$ uv run script.py # Automatically manages dependencies!
Virtual Environments
Create a virtual environment:
$ uv venv
$ uv venv .venv --python 3.12
Using pip-compatible interface:
$ uv pip install requests
$ uv pip install -r requirements.txt
$ uv pip freeze > requirements.txt
$ uv pip list
Python Version Management
Install Python versions:
$ uv python install 3.12
$ uv python install 3.11 3.10 # Multiple versions
List installed versions:
$ uv python list
Pin project Python version:
$ uv python pin 3.12
# Creates .python-version file
Tools Management
Install and run tools:
$ uv tool install ruff
$ uv tool install black mypy
$ uvx ruff check . # Run without installing
$ uvx pytest # Run tools in isolated environments
Building and Publishing
Build package:
$ uv build
# Creates dist/ with .whl and .tar.gz
Publish to PyPI:
$ uv publish
$ uv publish --token $PYPI_TOKEN
Configuration
Project config (pyproject.toml):
[project]
name = "my-app"
version = "0.1.0"
description = "My application"
readme = "README.md"
requires-python = ">=3.11"
dependencies = [
"requests>=2.31.0",
"pandas>=2.0.0",
]
[project.optional-dependencies]
dev = [
"pytest>=7.0.0",
"black>=23.0.0",
]
[dependency-groups]
dev = [
"ruff>=0.1.0",
]
[tool.uv.sources]
# Use alternative package indexes
my-package = { index = "private-pypi" }
Global config (uv.toml or pyproject.toml):
[tool.uv]
# Disable managed project environment
managed = false
# Configure package indexes
[[tool.uv.index]]
name = "pytorch"
url = "https://download.pytorch.org/whl/cu118"
Common Workflows
Workflow 1: Starting a New Project
# Create project
$ uv init my-project
$ cd my-project
# Add dependencies
$ uv add fastapi uvicorn sqlalchemy
# Add dev dependencies
$ uv add --dev pytest black ruff
# Run the app
$ uv run python main.py
Workflow 2: Migrating from pip/requirements.txt
# Create uv project from existing code
$ uv init --no-package
# Import existing requirements
$ uv add -r requirements.txt
# Lock dependencies
$ uv lock
# Sync environment
$ uv sync
Workflow 3: Docker Integration
FROM python:3.12-slim
# Install uv
COPY --from=ghcr.io/astral-sh/uv:latest /uv /bin/uv
# Copy project files
WORKDIR /app
COPY pyproject.toml uv.lock ./
COPY src ./src
# Install dependencies
RUN uv sync --frozen --no-cache
# Run application
CMD ["uv", "run", "python", "-m", "myapp"]
Workflow 4: GitHub Actions CI/CD
name: CI
on: [push, pull_request]
jobs:
test:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install uv
uses: astral-sh/setup-uv@v3
- name: Set up Python
run: uv python install 3.12
- name: Install dependencies
run: uv sync --all-extras --dev
- name: Run tests
run: uv run pytest
Workflow 5: Working with Scripts
# Create standalone script with dependencies
$ cat > analyze.py << 'EOF'
# /// script
# requires-python = ">=3.11"
# dependencies = [
# "pandas",
# "matplotlib",
# ]
# ///
import pandas as pd
import matplotlib.pyplot as plt
df = pd.read_csv("data.csv")
print(df.describe())
EOF
# Run with automatic dependency management
$ uv run analyze.py
Key Concepts
1. Projects
- Self-contained Python applications with
pyproject.toml - Automatic dependency locking with
uv.lock - Environment management with
uv sync
2. Workspaces
- Monorepo support for multiple related packages
- Shared dependency resolution
- Defined in root
pyproject.toml
3. Resolution
- Fast dependency resolver written in Rust
- Platform-specific resolution
- Supports complex version constraints
4. Cache
- Global package cache at
~/.cache/uv/(Linux/macOS) - Deduplication across projects
- Use
uv cache cleanto clear
5. Python Versions
- Automatic Python installation and management
.python-versionfile for project pinning- Multiple versions side-by-side
Performance Tips
- Use
uv sync --frozenin CI to skip lockfile updates - Leverage the cache - don't use
--no-cacheunless necessary - Use
uvxfor tools instead of installing globally - Pin Python versions with
.python-versionfor reproducibility - Use workspaces for monorepos to share dependencies
Integration Guides
PyTorch with CUDA
$ uv init pytorch-project
$ uv add torch --index pytorch --index-url https://download.pytorch.org/whl/cu118
Jupyter Notebooks
$ uv add jupyter ipykernel
$ uv run jupyter notebook
Pre-commit Hooks
# .pre-commit-config.yaml
repos:
- repo: https://github.com/astral-sh/uv-pre-commit
rev: 0.5.0
hooks:
- id: uv-lock
FastAPI Development
$ uv init --package my-api
$ uv add fastapi uvicorn[standard]
$ uv add --dev pytest httpx
$ uv run uvicorn myapi.main:app --reload
Reference Files
Comprehensive documentation in references/:
- uv.md - Complete UV documentation (63 pages)
- CLI reference for all commands
- Detailed guides for projects, packages, and tools
- Integration examples (Docker, GitHub, Jupyter, etc.)
- Advanced topics (authentication, indexes, build backends)
Troubleshooting
Common Issues
Issue: "No solution found"
- Check version constraints in
pyproject.toml - Try
uv lock --upgradeto get latest compatible versions - Use
uv treeto inspect dependency conflicts
Issue: "Python version not found"
- Install with
uv python install 3.12 - Check
.python-versionfile - Verify
requires-pythoninpyproject.toml
Issue: "Package not found"
- Verify package name spelling
- Check if package requires alternative index
- Use
--index-urlfor custom indexes
Migration Guide
From tool.uv.dev-dependencies (Deprecated)
IMPORTANT: The tool.uv.dev-dependencies field is deprecated and will be removed in a future release. Use [dependency-groups] instead (PEP 735).
# ❌ Old (deprecated)
[tool.uv]
dev-dependencies = [
"pytest>=7.0.0",
"ruff>=0.1.0",
]
# ✅ New (correct)
[dependency-groups]
dev = [
"pytest>=7.0.0",
"ruff>=0.1.0",
]
When you run uv add --dev pytest, it will automatically use the new [dependency-groups] format for new projects.
From pip + requirements.txt
# Before: pip install -r requirements.txt
# After:
$ uv init --no-package
$ uv add -r requirements.txt
$ uv sync
From Poetry
# Before: poetry install
# After:
$ uv init --no-package
$ uv add $(poetry show --only main | awk '{print $1}')
$ uv sync
From Pipenv
# Before: pipenv install
# After:
$ uv init --no-package
$ uv add $(pipenv requirements | grep -v "^#" | grep -v "^$")
$ uv sync
Best Practices
- Always commit
uv.lockfor reproducible builds - Use
pyproject.tomlinstead ofrequirements.txt - Pin Python version with
.python-version - Separate dev dependencies with
--devflag - Use workspaces for monorepos
- Leverage
uvxfor one-off tool execution - Configure indexes in
pyproject.tomlfor consistency
Resources
- Official Docs: https://docs.astral.sh/uv/
- GitHub: https://github.com/astral-sh/uv
- Installation:
curl -LsSf https://astral.sh/uv/install.sh | sh
Notes
- This skill was generated from official UV documentation (v0.5.0+)
- All code examples are tested and production-ready
- UV is actively developed - check docs for latest features
- UV is 10-100x faster than pip for most operations
04 同仓库更多 Skills
seedance-prompt-en
Write effective prompts for Jimeng Seedance 2.0 multimodal AI video generation. Use when users want to create video prompts using text, images, videos, and audio inputs with the @ reference system. Covers camera movements, effects replication, video extension, editing, music beat-matching, e-commerce ads, short dramas, and educational content.
onepage-pdf
Convert an HTML page into a single continuous-page PDF (one tall page, no pagination breaks), preserving desktop layout, backgrounds and selectable text. Use when the user asks to turn an HTML file/report/proposal into a "single-page PDF", "long PDF", "不分页 PDF", "单页 PDF", "长图式 PDF", or complains that an HTML-to-PDF export breaks into pages or cuts content. Supports optional string redaction with leak verification.
native-feel-cross-platform-desktop
Use when the user is designing, prototyping, or rewriting a desktop app that must run on multiple OSes (macOS + Windows, optionally Linux) AND feel indistinguishable from a native app to its users — fast launch, native windowing, native input handling, native materials. Trigger words include "cross-platform desktop", "Electron alternative", "Tauri vs native", "WebView wrapper", "near-native performance", "Raycast architecture", "WebKit/WebView2 quirks", "WKWebView", "system tray app", "global hotkey app", "launcher app". Do NOT trigger this skill for pure web apps, pure mobile apps, or for greenfield projects that have no native-feel requirement.
gstack
Router for the gstack skill suite. (gstack)
fireworks-tech-graph
Use when the user wants to create any technical diagram - architecture, data flow, flowchart, sequence, agent/memory, or concept map - and export as SVG+PNG. Trigger on: "画图" "帮我画" "生成图" "做个图" "架构图" "流程图" "可视化一下" "出图" "generate diagram" "draw diagram" "visualize" or any system/flow description the user wants illustrated.
dbs-video-extract
接收抖音、小红书、微信视频号等短视频链接或分享文案,查询作品/账号数据并提取语音文字稿,按作者和标题归档为 Markdown。用户希望完整解析短视频、取得数据或文字稿时使用;运行前检查 TikHub 与轻抖使用凭证,缺少时用新手能理解的方式引导购买和安全配置。