The Berkeley Pacman Projects repository offers a modular Python codebase where users build and test AI agents in a Pacman maze. It guides learners through uninformed and informed search (DFS, BFS, A*), adversarial multi-agent search (minimax, alpha-beta pruning), and reinforcement learning (Q-learning with feature extraction). Integrated graphical interfaces visualize agent behavior in real time, while built-in test cases and an autograder verify correctness. By iterating on algorithm implementations, users gain practical experience in state space exploration, heuristic design, adversarial reasoning, and reward-based learning within a unified game framework.

Mockaroni is a platform designed to generate custom synthetic text data that looks and feels similar to real-world data. The generated data can be used for various applications such as training AI and machine learning models, testing algorithms, and more. With customizable templates and advanced generation algorithms, Mockaroni ensures your models are well-prepared for real-world data scenarios, enhancing their efficiency and effectiveness.

ANAC-agents is a Python-based framework that centralizes multiple negotiation agent implementations for the Automated Negotiating Agents Competition (ANAC). Each agent within the repository embodies distinct strategies for utility modeling, proposal generation, concession tactics, and acceptance criteria, facilitating comparative studies and rapid prototyping. Users can define negotiation domains with custom issues and preference profiles, then simulate bilateral negotiations or tournament-style competitions across agents. The toolkit includes configuration scripts, evaluation metrics, and logging utilities to analyze negotiation dynamics. Researchers and developers can extend existing agents, test novel algorithms, or integrate external learning modules, accelerating innovation in automated bargaining and strategic decision-making under incomplete information.

Gym-Recsys is a toolbox that wraps recommendation tasks into OpenAI Gym environments, allowing reinforcement learning algorithms to interact with simulated user-item matrices step by step. It provides synthetic user behavior generators, supports loading popular datasets, and delivers standard recommendation metrics like Precision@K and NDCG. Users can customize reward functions, user models, and item pools to experiment with different RL-based recommendation strategies in a reproducible manner.

At its core, Gomoku Battle provides a robust simulation environment where AI agents adhere to a JSON-based protocol to receive board state updates and submit move decisions. Developers can integrate custom strategies by implementing simple Python interfaces, leveraging provided sample bots for reference. The built-in tournament manager automates scheduling of round-robin and elimination matches, while detailed logs capture metrics like win rates, move times, and game histories. Outputs can be exported as CSV or JSON for further statistical analysis. The framework supports parallel execution to accelerate large-scale experiments and can be extended to include custom rule variations or training pipelines, making it ideal for research, education, and competitive AI development.

Halite II is an open-source challenge framework that hosts turn-based strategy matches between user-written bots. Each turn, agents receive a map state, issue movement and attack commands, and compete to control the most territory. The platform includes a game server, map parser, and visualization tool. Developers can test locally, refine heuristics, optimize performance under time constraints, and submit to an online leaderboard. The system supports iterative bot improvements, multi-agent cooperation, and custom strategy research in a standardized environment.