Here’s a clean GitHub Release description for the first version (v1.0.0) you can paste directly into the release page:

🎉 WindMouse v1.0.0 — Initial Release

This is the first stable release of WindMouse, a Python library for generating human-like mouse movements using a physics-based WindMouse algorithm.

WindMouse is designed for automation scenarios where realistic cursor behavior matters, helping avoid detection by producing non-linear paths, variable speed, and natural deceleration.

✨ Key Features

• Physics-based WindMouse algorithm

  • Gravity-driven target attraction
  • Randomized wind force for natural curvature
  • Smooth acceleration and deceleration near the target

• Human-like mouse movement

  • Curved, non-deterministic trajectories
  • Variable speed instead of constant linear motion

• Multiple backends

  • PyAutoGUI (cross-platform: Windows, macOS, Linux)
  • AutoHotkey (Windows-only, native integration)

• Fine-grained control

  • Configurable gravity, wind, speed, and damping parameters
  • Adjustable movement timing (tick_delay, step_duration)

• Advanced interaction support

  • Drag & drop via mouse button holding
  • Dynamic destination changes during movement
  • Manual step-by-step control with tick()

• Type-safe API

  • Strong typing with NewType (Coordinate)
  • Fully compatible with mypy

• Well-tested & documented

  • Comprehensive unit and integration tests
  • Detailed documentation covering algorithm theory, API, and usage examples

📦 What’s Included

• Core WindMouse path generator
• Abstract controller interface
• PyAutoGUI and AutoHotkey controllers
• Sphinx-based documentation
• CI pipelines for testing, linting, and publishing

🚀 Getting Started

pip install windmouse[pyautogui]

from windmouse.pyautogui_controller import PyautoguiMouseController
from windmouse import Coordinate

mouse = PyautoguiMouseController()
mouse.dest_position = (Coordinate(800), Coordinate(600))
mouse.move_to_target()

🔖 Versioning

This release marks v1.0.0, establishing a stable public API.
Future releases may introduce new movement models, optimizations, and additional backends while maintaining backward compatibility when possible.

Feedback, bug reports, and contributions are welcome! 🚀