Skip to content

PG-AGI/PyRefine

Repository files navigation

PyRefine

1. Download the latest binary

The prebuilt executables live on the GitHub Releases page. Replace v1.0 with the newest tag when needed.

Windows

curl -L -o pyrefine.exe https://github.com/PG-AGI/PyRefine/releases/download/v1.0/pyrefine.exe
pyrefine.exe --version

macOS

curl -L -o pyrefine-macos https://github.com/PG-AGI/PyRefine/releases/download/v1.0/pyrefine-macos
chmod +x pyrefine-macos

./pyrefine-macos --version

Ubuntu/Linux

curl -L -o pyrefine-linux https://github.com/PG-AGI/PyRefine/releases/download/v1.0/pyrefine-linux
chmod +x pyrefine-linux

./pyrefine-linux --version

If the executable is stored outside the project root, pass --project-root PATH_TO_REPO on every command (e.g. pyrefine.exe --project-root C:\code\MyRepo --setup).

2. Standard workflows

A. Bootstrap a brand-new project

  1. Download the appropriate binary and place it in your empty project directory.
  2. Run the initializers:
    pyrefine --create
    
    pyrefine --setup
    
    pyrefine --clean .
    (On Windows use pyrefine.exe.)
  3. Commit the generated structure and begin development.

B. Clean up or adopt an existing project

  1. Drop the binary into the repository (or reference it via --project-root).
  2. Apply the cleanup workflow:
    pyrefine --setup
    
    pyrefine --clean .
    
    pyrefine --test-coverage   # optional validation

3. Commands

Command Summary
pyrefine.exe Runs clean workflow on current repo
pyrefine --clean <path> Formats target using full formatter stack
pyrefine --create Bootstraps standard project structure
pyrefine --setup Configures VS Code and environments
pyrefine --test-coverage [path] Generates pytest coverage reports per project
pyrefine --update [--manifest-url URL] Downloads and installs newest binary release

4. How PyRefine works (detailed)

If you want the complete working details for developers and maintainers, see docs/WORKING.md which describes the architecture, each command's behavior, and implementation notes.

About

To develop an intelligent Python code quality assistant that automatically detects, highlights, and fixes linting errors, enforces best coding practices, and enhances overall code readability, maintainability, and reliability, enabling developers to write clean, efficient, and Pythonic code with minimal manual intervention.

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Languages