Project Library

Fluent Bit is a high-performance, lightweight log, metrics, and traces processor and forwarder, part of the CNCF and Fluentd ecosystem. It supports Linux, Windows, MacOS, and BSD systems, running on x86, ARM, and other architectures. Key features include low CPU/memory usage, data parsing (JSON, Regex, LTSV, Logfmt), reliability with backpressure handling and buffering, and built-in TLS/SSL for secure networking. It offers a pluggable architecture with over 70 built-in plugins for inputs, filters, and outputs, extensible in C, Lua, and Golang. Fluent Bit also provides SQL-based stream processing for data manipulation, analytics, and timeseries forecasting. Widely adopted in production, it integrates with backends like Elasticsearch, Kafka, AWS, Azure, and more, making it a versatile solution for log and metrics processing.

**DeskHop** is a fast, open-source desktop switching solution designed for users managing multiple computers with a single keyboard and mouse. It enables seamless switching between computers, even across different operating systems (Linux, macOS, Windows), by dragging the mouse pointer or using a keyboard shortcut. Key features include no noticeable delay, full galvanic isolation, and no need for installed software. The project is affordable, with components costing under €15, and includes a 3D printable case. DeskHop supports advanced functionalities like mouse slowdown, switch lock, gaming mode, and screensaver. It is built using Raspberry Pi Pico boards, ensuring cost-effectiveness and ease of assembly. The project is entirely free, open-source, and prioritizes security by preventing data leakage between systems.

Opik is an open-source framework designed for evaluating, testing, and monitoring Large Language Model (LLM) applications. It supports a wide range of use cases, including RAG chatbots, code assistants, and complex agentic pipelines. Key features include tracing LLM calls, annotating feedback, and automating evaluations using LLM-as-a-judge metrics for tasks like hallucination detection and RAG evaluation. Opik integrates with popular frameworks like OpenAI, LangChain, and LlamaIndex, and offers CI/CD pipeline compatibility. It provides robust production monitoring with dashboards and online evaluation metrics. Available as a hosted solution or for local deployment, Opik aims to enhance LLM system performance, speed, and cost-efficiency.

Mermaid is a JavaScript-based diagramming tool that generates diagrams and charts from Markdown-like text definitions. It simplifies the creation of complex diagrams, making it accessible even to non-programmers through its Live Editor. Mermaid supports a wide range of diagram types, including flowcharts, sequence diagrams, Gantt charts, class diagrams, state diagrams, pie charts, and more. It integrates seamlessly with platforms like GitHub and can be embedded in various applications. Designed to keep documentation up-to-date with development, Mermaid reduces the time and effort required to maintain accurate diagrams. Its text-based syntax allows for easy modifications, and it offers features like sandboxed rendering for enhanced security. Mermaid is open-source, community-driven, and widely used for visualizing processes, systems, and data.

ConfettiSwiftUI is a lightweight, customizable SwiftUI library for creating dynamic confetti animations in iOS and macOS applications. Built entirely with SwiftUI, it allows developers to trigger animations with a single state change, offering haptic feedback for each explosion. Users can choose from default shapes, emojis, SF Symbols, or custom text for confetti elements. The library supports extensive customization, including color, size, repetition, and explosion radius, and integrates seamlessly via Swift Package Manager or manual installation. Ideal for celebratory effects in apps, it is compatible with iOS 14.0+, macOS 11+, and Swift 5+. A configurator app is available for live preview and fine-tuning animations.

Grip is a Python-based command-line tool that renders local Markdown files using GitHub's Markdown API, ensuring the preview matches GitHub's exact styling. It serves as a local server, allowing instant browser updates without page refreshes. Key features include exporting to HTML, support for user-content rendering, and customizable configurations like port and host settings. Grip also offers API access for integration into other projects and supports authentication to bypass GitHub's rate limits. Ideal for Readme-driven development, it provides a seamless way to preview and refine documentation before pushing to GitHub.

Lima is a versatile tool for launching Linux virtual machines with automatic file sharing and port forwarding, designed to simplify Linux environments on macOS, Linux, and NetBSD. Initially focused on promoting containerd and nerdctl for Mac users, Lima supports various container engines like Docker, Podman, and Kubernetes, as well as non-container applications. It offers seamless integration with commands for Linux operations and container management, making it ideal for developers and system administrators. Lima is adopted by tools like Rancher Desktop, Colima, Finch, and Podman Desktop, and features community-driven GUIs and plugins. As a Cloud Native Computing Foundation sandbox project, Lima emphasizes ease of use, flexibility, and community collaboration.

Letta is an open-source framework designed for building stateful, long-term memory-enabled LLM applications. It allows developers to create advanced reasoning agents with transparent memory management, supporting various LLM backends like OpenAI, Anthropic, and Ollama. Letta offers a white-box, model-agnostic approach, enabling users to deploy agents via Docker or pip, with PostgreSQL as the recommended database for persistence. The framework includes the Letta Agent Development Environment (ADE), a graphical interface for managing and interacting with agents, and provides REST API and SDKs for integration. Letta is ideal for applications requiring persistent, intelligent agents, such as customer support chatbots, and supports both self-hosted and cloud-based deployments.

LaTeX.CSS is a minimal, class-less CSS library designed to give websites the appearance of a LaTeX document. It is lightweight and easy to integrate via a `<link>` tag, CDN, or NPM/Yarn installation. The library supports optional class-based elements for specialized styles like theorems, lemmas, and abstracts, and offers multilingual support for labels in various languages. Built on David Zollikofer's latexCSS, it incorporates modern CSS reset principles and features inspired by Tufte CSS. LaTeX.CSS is open-source under the MIT License, encouraging contributions and feedback for continuous improvement.

This project is a hands-on tutorial for implementing an open-addressed, double-hashed hash table in C. It provides a clear, step-by-step guide to understanding the inner workings of hash tables, including their structure, hash functions, collision handling, and resizing. The tutorial is designed for those with basic programming and C knowledge, offering practical exposure to low-level coding and data structure concepts. With around 200 lines of code, it takes approximately 1-2 hours to complete, making it an efficient way to deepen your understanding of hash tables and their applications in computer science.

LÖVE is a free, open-source framework for creating 2D games using Lua. It supports multiple platforms, including Windows, macOS, Linux, Android, and iOS. The framework provides comprehensive documentation via its wiki, active community support through forums, Discord, and Reddit, and a robust test suite for API validation. Development occurs on the 'main' branch, with stable releases tagged and available for download. Nightly builds are also accessible for testing. Contributions are encouraged via pull requests, with discussions on new features and changes taking place on Discord or the issue tracker. LÖVE relies on dependencies like SDL3, OpenGL, Lua, and FreeType, and offers detailed build instructions for various platforms, including Windows, macOS, iOS, and Android.

QTTabBar is a lightweight tool designed to enhance Windows Explorer with tabbed browsing, similar to web browsers like Firefox or Opera. It allows users to manage multiple folders efficiently through tabs, improving productivity. The project includes features such as file operation tools, tree directories, and a status bar. It supports customization, including dark mode and background images, and is optimized for Chinese users with localized features. Compatible with Windows 11, it requires .NET Framework 3.5. The tool is regularly updated, with recent versions adding multi-language support, debugging logs, and DPI adjustments. QTTabBar is available on GitHub, Gitee, and SourceForge, with detailed installation and usage guides provided.

Bocker is a minimalist implementation of Docker, written in approximately 100 lines of bash. It replicates core Docker functionalities such as image management (`pull`, `images`, `commit`), container operations (`run`, `exec`, `rm`), and basic networking. Bocker also supports resource management via CGroups for CPU and memory limits. Designed for simplicity, it requires specific system configurations, including a Btrfs filesystem and a network bridge. Due to its low-level operations and root access, it is recommended to run Bocker in a virtual machine. While it lacks advanced features like data volumes and port forwarding, Bocker serves as a lightweight, educational tool for understanding containerization principles. Licensed under GNU GPLv3, it is free to use and modify.

Jitsi Meet is an open-source video conferencing platform offering high-quality HD audio and video, accessible via browsers and mobile apps. Key features include content sharing, chat with private conversations, polls, virtual backgrounds, and interactive tools like raise hand and reactions. It supports web and native SDKs for seamless integration and is scalable for both personal and enterprise use. Users can start meetings instantly with a Google, Facebook, or GitHub account or deploy their own instance using Debian packages or Docker. For enterprise needs, Jitsi as a Service (JaaS) provides a managed solution with branding capabilities. The platform emphasizes security with end-to-end encryption and comprehensive documentation for developers and contributors.

`runlike` is a Docker utility that generates the command-line instructions needed to replicate an existing container, including all configurations such as ports, volumes, and environment variables. It simplifies the process of manually recreating containers, especially for users deploying via configuration management tools like Ansible or Chef. The tool can be used directly by providing a container name or by processing `docker inspect` output. It supports common Docker run options and offers features like pretty-printing commands and omitting container names to avoid conflicts. Available as a Docker image or via `pip install`, `runlike` is a time-saving solution for container management, though it is still under development and not recommended for production use without verification.