Project Library

Zap is a high-performance, structured, and leveled logging library for Go, designed for applications where logging speed and efficiency are critical. It offers two logging options: the `SugaredLogger` for ease of use with both structured and `printf`-style APIs, and the `Logger` for maximum performance and type safety with structured logging only. Zap achieves its speed by avoiding reflection and minimizing allocations, making it significantly faster than many other logging libraries, including the Go standard library. It supports the two most recent minor versions of Go and is stable, with no breaking changes expected in its 1.x series. Ideal for performance-sensitive applications, Zap provides a flexible and efficient logging solution.

**zoxide** is a smarter, faster alternative to the `cd` command, designed to enhance directory navigation by learning and ranking frequently accessed directories. Inspired by tools like `z` and `autojump`, it allows users to "jump" to directories with minimal keystrokes. Compatible with all major shells (Bash, Zsh, Fish, PowerShell, etc.), zoxide supports interactive selection via `fzf` and offers customizable commands and configurations. It integrates seamlessly with various file managers, text editors, and terminal tools, making it a versatile solution for efficient directory management. Installation is straightforward across multiple platforms, and it supports data import from similar tools like `autojump` and `z`.

Kestra is an open-source, event-driven orchestration platform designed for both scheduled and real-time workflows. It leverages Infrastructure as Code principles, allowing users to define workflows in YAML directly from an intuitive UI or code editor. Key features include Git version control integration, a rich plugin ecosystem for diverse integrations, and support for scripting in multiple languages. Kestra ensures scalability, fault tolerance, and high availability, making it suitable for handling millions of workflows. It offers advanced workflow management with features like namespaces, labels, subflows, retries, and dynamic tasks. The platform is developer-friendly, supporting CI/CD pipelines, Terraform, and custom plugin development. Kestra’s UI provides real-time validation, auto-completion, and a visual DAG representation for seamless workflow creation and monitoring.

MagicMirror is an AI-powered desktop application that enables instant face swapping, hairstyle, and outfit transformations with a single click. Designed for simplicity, it allows users to drag and drop photos for instant changes without complex settings. The app is lightweight, with a <10MB installer and <1GB model files, and runs efficiently on standard computers without requiring dedicated GPU hardware. It prioritizes privacy by processing all data offline, ensuring images never leave the user’s device. MagicMirror supports macOS 13 (Ventura) and Windows 10 and above, making it accessible for personal entertainment and creative exploration. Built on open-source technologies like Tauri and InsightFace, it emphasizes ethical usage and prohibits commercial applications.

The Generative AI for Krita plugin integrates advanced AI image generation tools directly into Krita, enhancing painting and editing workflows. Key features include inpainting, live painting, upscaling, and support for models like Stable Diffusion and ControlNet. It offers precision and control through region-specific text prompts, reference images, and various control layers. The plugin emphasizes seamless workflow integration, allowing users to draw, paint, and generate images without technical interruptions. It supports local, open-source models, with options for cloud generation, and is compatible with Windows, Linux, and macOS. Customization is extensive, enabling users to create presets and use their own models. Installation is straightforward, requiring Krita 5.2.0 or newer, and it supports powerful GPUs for optimal performance.

**semantic-release** is a fully automated tool for version management and package publishing, designed to streamline the release workflow. It enforces Semantic Versioning (SemVer) by analyzing commit messages to determine the next version number, generate release notes, and publish packages. Key features include integration with CI workflows, support for multiple distribution channels, and compatibility with various package managers and languages via plugins. It eliminates manual errors, ensures immediate availability of new features and fixes, and notifies users of updates. The tool relies on formalized commit message conventions and can be configured for custom workflows, promoting consistency and efficiency in software releases.

**garak** is a free, open-source vulnerability scanner designed to test Large Language Models (LLMs) for weaknesses such as hallucination, data leakage, prompt injection, misinformation, toxicity, and jailbreaks. It employs static, dynamic, and adaptive probes to identify failure modes in LLMs, similar to tools like `nmap` or Metasploit but tailored for generative AI. Supporting a wide range of LLMs (e.g., Hugging Face, OpenAI, Replicate), garak is a command-line tool that can be installed via `pip` or cloned from GitHub. It provides detailed logging, customizable probes, and detectors, making it a versatile tool for red-teaming and assessing LLM security. Developed by NVIDIA, garak is actively maintained and welcomes community contributions.

The Micro Racer Car project is an open-source, ESP32-based miniature car featuring firmware, hardware, mechanical designs, and neural network scripts. It includes detailed schematics, PCB gerber files, and 3D-printable enclosures and wheels. The car can be controlled via a custom remote or keyboard, and users can train a neural network for advanced functionalities. The project encourages further development, such as integrating color sensors for line-following or LLM-based control. Licensed under Apache 2.0, it’s designed for hobbyists and developers to expand and customize. Contributions and improvements are welcomed, fostering a collaborative open-source community.

Perplexica is an open-source, AI-powered search engine designed to deliver precise, up-to-date answers by leveraging advanced machine learning techniques like similarity searching and embeddings. It integrates with SearxNG to ensure privacy and real-time information retrieval. Key features include support for local LLMs (e.g., Llama3, Mixtral via Ollama), two search modes (Normal and Copilot), and six focus modes tailored for specific needs like academic research, YouTube, Reddit, and Wolfram Alpha queries. Perplexica offers an API for seamless integration into applications and supports Docker for easy installation. It prioritizes user privacy and provides cited sources for transparency, making it a versatile tool for enhanced web searching.

Hertz is a high-performance, high-usability, and high-extensibility Golang HTTP framework designed for building microservices. It integrates a self-developed high-performance network library, Netpoll, and supports on-demand switching to Go Net. Hertz offers multi-protocol support, including HTTP/1.1 and ALPN, with customizable protocol resolution. Its layered design ensures extensive extensibility, allowing users to tailor the framework to their needs. Widely used within ByteDance, Hertz is ideal for developers prioritizing microservice performance and customization. The framework also provides a rich ecosystem of extensions for features like HTTP/2, WebSocket, monitoring, and more, making it a versatile choice for modern microservice architectures.

TRELLIS is a state-of-the-art 3D asset generation model that produces high-quality 3D outputs from text or image inputs. It leverages a unified Structured Latent (SLAT) representation, enabling versatile outputs such as Radiance Fields, 3D Gaussians, and meshes. With up to 2 billion parameters trained on a diverse dataset of 500K 3D objects, TRELLIS outperforms existing methods in quality and flexibility. Key features include intricate shape and texture details, support for multiple 3D formats, and capabilities for local editing and generating variants of objects. It also offers pre-trained models, a large-scale dataset, and tools for training and fine-tuning, making it a powerful solution for scalable and adaptable 3D generation.

PDFMathTranslate is a versatile tool designed for translating scientific PDF documents while preserving their original structure, including formulas, charts, tables of contents, and annotations. It supports multiple languages and integrates various translation services. The project offers multiple usage options, including a command-line tool, an interactive GUI, and Docker deployment, making it accessible for diverse user needs. It also provides online demos for quick testing and supports advanced features like partial document translation, multi-threading, and custom prompts. PDFMathTranslate is continuously updated with experimental backends and improved functionalities, ensuring flexibility and efficiency for academic and professional use.

Wave Terminal is an open-source, cross-platform terminal that integrates traditional command-line functionality with advanced graphical tools like file previews, web browsing, and AI assistance. Designed for modern development workflows, it eliminates the need to switch between terminals and browsers by embedding visual interfaces directly into the terminal. Key features include a drag-and-drop interface, a built-in editor with syntax highlighting, rich file previews, integrated AI chat with multiple model support, and one-click remote connections. It also offers customizable themes, terminal styles, and a powerful `wsh` command system for workspace management. Compatible with macOS, Linux, and Windows, Wave Terminal enhances productivity by keeping developers in their terminal environment while providing access to essential graphical tools.

Keep is an open-source AIOps and alert management platform designed to streamline incident response and monitoring. It offers a unified interface for alert deduplication, enrichment, filtering, and correlation, along with bi-directional integrations with various observability tools. Key features include customizable workflows, AI-powered incident summarization, and automation capabilities akin to GitHub Actions for monitoring. Keep supports a wide range of integrations, including AI backends, observability tools, databases, and communication platforms. It is enterprise-ready, offering robust security, flexible deployment options, and scalability for production environments. Developers can easily get started with Docker, Kubernetes, or cloud deployments, making it a versatile solution for modern IT operations.

Lutris is a versatile gaming platform designed to install and play video games across various eras and systems by integrating emulators, engine re-implementations, and compatibility layers. It provides a centralized interface to manage and launch games, connecting with services like Humble Bundle, GOG, and Steam for seamless library access. Game installations are automated using customizable JSON or YAML scripts. Lutris supports optional account creation for syncing game libraries and offers extensive configuration options for clients, runners, and games. It includes command-line tools for installation, debugging, and game management. Lutris is community-supported, encouraging contributions and donations to sustain development.