Automated Reinforcement Learning-Based Test Case Generation and Prioritization for Games

This project implements an automated framework for generating and prioritizing test cases for games using reinforcement learning (RL), static analysis, and model-based testing techniques. The system is designed to train an RL agent in a game-like environment and use its learned behavior to create intelligent and diverse test cases, followed by a code-coverage-based prioritization system.

🧠 Project Overview

The project consists of the following major components:

• Custom Game Environment: A TaxiGymEnv simulating a grid-based taxi game using Gymnasium.
• Reinforcement Learning Agent: Trained with PPO (Proximal Policy Optimization) using Stable-Baselines3 and Shimmy for compatibility.
• Finite State Machine (FSM): Encodes the game's logic and state transitions.
• Test Case Generator: Generates valid and partial behavior traces from the trained agent.
• Coverage Mapper: Uses coverage.py and ast to map test executions to the functions they touch.
• Prioritizer: Uses Git diff and coverage data to re-order test cases by relevance after code changes.

🧪 Requirements

• Python 3.10+
• Gymnasium
• Shimmy
• Stable-Baselines3
• Coverage.py
• Git (for diff-based prioritization)
• statemachine

Install requirements:

pip install gymnasium shimmy stable-baselines3 coverage gitpython statemachine