This open-source project delivers a complete multi-agent reinforcement learning framework built on PyTorch and Unity ML-Agents. It offers decentralized DDPG algorithms, environment wrappers, and training scripts. Users can configure agent policies, critic networks, replay buffers, and parallel training workers. Logging hooks allow TensorBoard monitoring, while modular code supports custom reward functions and environment parameters. The repository includes sample Unity scenes demonstrating collaborative navigation tasks, making it ideal for extending and benchmarking multi-agent scenarios in simulation.

Swarm Simulator provides a customizable environment for real-time multi-agent experiments. Users can adjust key behavior parameters—alignment, cohesion, separation—and observe emergent dynamics on a visual canvas. It supports interactive UI sliders, dynamic agent count adjustment, and data export for analysis. Ideal for educational demonstrations, research prototyping, or hobbyist exploration of swarm intelligence principles.

Arcade Vercel AI Template is an open-source boilerplate designed to kickstart AI-powered web projects using Vercel’s AI SDK. It provides pre-built components for chat interfaces, serverless API routes, and agent configuration files. Through a simple file structure, developers define their AI agents, prompts, and model parameters. The template handles authentication, routing, and deployment settings out of the box, enabling rapid iteration. By leveraging ArcadeAI’s APIs, users can integrate generative text, database lookups, and custom business logic. The result is a scalable, maintainable AI website that can be deployed in minutes to Vercel’s edge network.

PeerReview is a Chrome extension designed to summarize any highlighted text instantly. Utilizing Gemini's Prompt API and Summarizer API, it offers a practical solution for users who need quick text summaries. This tool is particularly useful for students, researchers, and professionals who often deal with large volumes of text and need a way to condense information rapidly. As an open-source project, PeerReview also welcomes contributions from developers looking to improve its functionality.

The Reddit News Agent System leverages the Multi-Channel Pipeline (MCP) for modular data processing and the Agent Development Kit (ADK) for workflow orchestration. After configuration, it continuously monitors chosen subreddits, applies sentiment analysis, topic classification, and summary generation modules, then routes the results to email, messaging apps, or dashboard interfaces. Developers can extend pipelines with custom processors, integrate new delivery channels, and fine-tune agent behaviors for tailored news curation and automated reporting.

AI Agents for Rock Paper Scissors is an open-source Python project that demonstrates how to build, train, and evaluate different AI strategies—random play, rule-based pattern recognition, and reinforcement learning (Q-learning)—in the classic Rock-Paper-Scissors game. It provides modular agent classes, a configurable game runner, performance logging, and visualization utilities. Users can easily swap agents, adjust learning parameters, and explore AI behavior in competitive scenarios.

HexHoot is an open-source project designed to create a communication platform that respects user privacy and data ownership. By eliminating the need for centralized servers, it ensures that all data is stored locally on users' devices. HexHoot uses advanced zero-knowledge-proof strategies to authenticate users without compromising security. This approach makes it ideal for secure, transparent, and decentralized communication, free from the risks of traditional P2P software dependencies.