Github Top Repositories

🎯 codecrafters-io/build-your-own-x landed on trending. Worth a proper look. 🔗 https://github.com/codecrafters-io/build-your-own-x 📝 Master programming by recreating your favorite technologies from scratch. ────────────────────────────── The codecrafters-io/build-your-own-x GitHub repository is a treasure trove of guides for building various technologies from scratch. Its purpose is to help you learn by creating real-world projects, following the philosophy of Richard Feynman: "What I cannot create, I do not understand." The repo offers a wide range of key features, including step-by-step tutorials, example code in multiple programming languages, and explanations of complex concepts. You can use these guides to build your own 3D renderer, AI model, augmented reality app, BitTorrent client, blockchain, bot, command-line tool, database, and many more. Some technical highlights include the use of languages like C++, Java, Python, and JavaScript, as well as frameworks and libraries such as Unity and OpenCV. The guides also cover various topics, such as computer graphics, machine learning, networking, and cybersecurity. This repository is suitable for developers of all levels, from beginners to experienced professionals, who want to improve their skills and learn by building real-world projects. In short, building your own technology from scratch is the best way to learn, and this repository provides the perfect starting point - so, get building and create something amazing! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

📌 Spotted on GitHub Trending: iv-org/invidious — let's break it down. 🔗 https://github.com/iv-org/invidious 📝 Invidious is an alternative front-end to YouTube ────────────────────────────── The Invidious project is an open-source alternative front-end to YouTube, offering a lightweight and ad-free experience. Its key features include no tracking, customizable homepage, and notifications for subscribed channels. Invidious can be used by selecting a public instance from the list or by hosting it yourself. From a technical standpoint, Invidious has embedded video support and a developer API, and it does not use official YouTube APIs. The project is available in many languages and has a strong focus on community involvement, with opportunities to contribute code and translate the platform. The target audience for Invidious includes anyone looking for a private and ad-free YouTube experience. Whether you're a casual user or a developer, Invidious provides a unique alternative to the traditional YouTube platform. Invidious is all about taking back control of your YouTube experience - it's your videos, your way. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🎯 lyogavin/airllm landed on trending. Worth a proper look. 🔗 https://github.com/lyogavin/airllm 📝 AirLLM 70B inference with single 4GB GPU ────────────────────────────── AirLLM is a game-changer for large language models, allowing them to run on significantly less GPU memory without sacrificing performance. With AirLLM, you can run massive models like 405B Llama 3.1 on just 8GB of GPU memory or DeepSeek-V3 (671B) on around 12GB. The secret lies in its ability to stream one expert at a time, rather than loading the entire layer. To get started, simply install the airllm package using pip install airllm and initialize the model with AutoModel.from_pretrained("model_id"). You can also enable model compression for up to 3x inference speedup by passing the compression argument. AirLLM supports a wide range of models, including ChatGLM, QWen, Baichuan, and more. It's perfect for researchers, developers, and enthusiasts looking to push the boundaries of language modeling. In short, AirLLM is a powerful tool that makes large language models more accessible and efficient. With its ease of use and impressive performance, it's a must-try for anyone working with language models: run bigger models, faster, and on less hardware. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

📌 Spotted on GitHub Trending: usekaneo/kaneo — let's break it down. 🔗 https://github.com/usekaneo/kaneo 📝 🎯 All you need. Nothing you don't. Open source project management that works for you, not against you. ────────────────────────────── Kaneo is a minimalist project management platform designed to help teams focus on building great products without unnecessary distractions. Its clean interface, self-hosted nature, and fast performance make it an attractive alternative to bloated tools. To get started with Kaneo, you can use the drim CLI tool for a one-click deployment or set it up with Docker Compose for a quick start. The platform is open source with a permissive MIT license, and its documentation provides detailed guides for setup, configuration, and development. Kaneo is suitable for teams looking for a simple, efficient, and customizable project management solution. Whether you're a developer, a project manager, or a team lead, Kaneo's flexible and adaptable nature makes it a great choice. From a technical standpoint, Kaneo's API and web services can be configured and customized to meet specific needs, and its Kubernetes deployment options make it easy to scale and manage. In short, Kaneo is a breath of fresh air in the world of project management tools - simple, fast, and efficient. Try it out and see the difference for yourself: with Kaneo, less is more, and that's what makes it great. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🌟 microsoft/AI-For-Beginners caught my eye on GitHub Trending today. 🔗 https://github.com/microsoft/AI-For-Beginners 📝 12 Weeks, 24 Lessons, AI for All! ────────────────────────────── The Microsoft AI-For-Beginners GitHub repository is an excellent starting point for those new to Artificial Intelligence (AI). This 12-week, 24-lesson curriculum is designed to be beginner-friendly and covers a wide range of topics, including TensorFlow and PyTorch, as well as ethics in AI. The curriculum is divided into sections, including an Introduction to AI, Symbolic AI, and Introduction to Neural Networks. Each section includes practical lessons, quizzes, and labs to help reinforce learning. One of the key features of this repository is its multi-language support, with translations available in over 50 languages. To get started, you can clone the repository locally, and to make the process faster, you can use sparse checkout to exclude translations. Here's an example of how to do this:

git clone --filter=blob:none --sparse https://github.com/microsoft/AI-For-Beginners.git
cd AI-For-Beginners
git sparse-checkout set --no-cone '/*' '!translations' '!translated_images'

The repository also includes a mindmap of the course to help visualize the topics covered. You can join the community and get involved through the Microsoft Foundry Discord server. Overall, the Microsoft AI-For-Beginners repository is a comprehensive resource for anyone looking to learn about AI. So, dive in and start exploring - with this curriculum, you'll be well on your way to becoming an AI master in no time! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🚀 Meet bytedance/deer-flow: a gem from today's GitHub trending list. 🔗 https://github.com/bytedance/deer-flow 📝 An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours. ────────────────────────────── DeerFlow is an open-source super agent harness that lets you orchestrate sub-agents, memory, and sandboxes to do almost anything, powered by extensible skills. Its key features include skills and tools, session goals, sub-agents, and context engineering. To get started, you can use the make setup command to run the setup wizard, which guides you through choosing an LLM provider and configuring your setup. You can also use the make dev command for local development or make docker-start for Docker development. For a quick start, you can clone the repository, run the setup wizard, and then start the application using make up. DeerFlow 2.0 is a ground-up rewrite with no shared code with version 1, and it's designed to be extensible and customizable. The project has a official website with real demos and a sister project called LLM Space. The recommended models include GPT-4o, Gemini 2.5 Flash, and Qwen3 32B. You can also use the embedded Python client to interact with DeerFlow. One-line takeaway: With DeerFlow, you can build a super agent harness that can do almost anything, and it's free and open-source. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🔥 NomaDamas/k-skill is trending — and it deserves your attention. 🔗 https://github.com/NomaDamas/k-skill 📝 한국인을 위한 스킬 모음집 - 에이전트를 한국인으로 ────────────────────────────── Hey, junior dev! Let's explore the k-skill GitHub repo. This project is a collection of skills for automating various tasks, mainly focused on Korean services. With k-skill, you can automate tasks like SRT and KTX booking, lottery checks, and even interactions with government services like tax and business registration. The key features of k-skill include support for multiple coding agents like Claude Code and Codex, as well as a simple installation process using npx. You can install all skills at once or choose specific ones to install. Some of the skills available in k-skill include: - SRT and KTX booking - Lottery checks - Interactions with government services like tax and business registration - Automated tasks for Korean services like banking and insurance The technical highlights of k-skill include its support for multiple coding agents and its use of Node.js for installation and execution. This project is aimed at developers who want to automate tasks related to Korean services. If you're interested in contributing to k-skill, be sure to check out the CONTRIBUTING.md file for more information. In short, k-skill is an amazing resource for automating tasks related to Korean services - and with great power comes great automation! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🔍 Deep-diving into TencentCloud/TencentDB-Agent-Memory — fresh off the trending list. 🔗 https://github.com/TencentCloud/TencentDB-Agent-Memory 📝 TencentDB Agent Memory is a team-level memory hub for AI Agents — turning conversations, docs, and code into four reusable memory assets (Chat Memory, Skill, LLM-Wiki, Code-Graph) that are governed, shared, and equipped across agents and frameworks. ────────────────────────────── The TencentDB Agent Memory is designed to help agents learn from workflows, retain task context, and reuse past experiences. It achieves this through a layered system consisting of symbolic memory for in-task information and memory layering for cross-session experience. Key features include memory layering with heterogeneous storage, symbolic memory using Mermaid syntax, and context offloading with `node_id` tracing. This approach enables agents to reason better, not just remember more. Technical highlights include a unified architectural paradigm with short-term context layering, long-term personalization layering, and skill generation layering. The system also ensures full traceability and lossless recovery through a deterministic path from high-level abstractions to ground-truth evidence. The target audience includes developers and users of OpenClaw and Hermes agents, who can integrate this plugin to enhance their agents' memory capabilities. To get started, users can follow the quick start guide, which includes installation and configuration instructions for OpenClaw and Hermes. In summary, the TencentDB Agent Memory plugin is a powerful tool for improving agent performance by enabling them to learn from experience and retain context. Let the agents remember, so humans can innovate. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

💡 microsoft/TRELLIS.2 just hit the trending charts — here's why it matters. 🔗 https://github.com/microsoft/TRELLIS.2 📝 Native and Compact Structured Latents for 3D Generation ────────────────────────────── TRELLIS.2 is a state-of-the-art 3D generative model that enables high-fidelity image-to-3D generation. It features a novel "field-free" sparse voxel structure called O-Voxel, which allows for the reconstruction and generation of complex 3D assets with sharp features and full PBR materials. The model boasts high-quality, resolution, and efficiency, generating high-resolution fully textured assets with exceptional fidelity. Some of the key features of TRELLIS.2 include: - Arbitrary topology handling: The O-Voxel representation can handle complex structures without lossy conversion, including open surfaces, non-manifold geometry, and internal enclosed structures. - Rich texture modeling: The model supports arbitrary surface attributes, including base color, roughness, metallic, and opacity, enabling photorealistic rendering and transparency support. - Minimalist processing: Data processing is streamlined for instant conversions that are fully rendering-free and optimization-free. To use TRELLIS.2, you'll need to install the dependencies, including the CUDA Toolkit and Conda, and then clone the repository. You can then use the pretrained model for image-to-3D generation or PBR texture generation. The repository also provides a web demo for easy testing. From a technical standpoint, TRELLIS.2 is a 4B-parameter model that utilizes a Sparse 3D VAE with 16× spatial downsampling to encode assets into a compact latent space. The model is designed for high-performance and can generate high-resolution assets with exceptional fidelity. The target audience for TRELLIS.2 includes researchers, developers, and artists working with 3D generation and image-to-3D applications. With its powerful features and streamlined processing, TRELLIS.2 is an ideal choice for anyone looking to push the boundaries of 3D generation. In short, TRELLIS.2 is a game-changer for 3D generation, offering unparalleled fidelity, efficiency, and flexibility - and with its open-source availability, the possibilities are endless! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🌟 ansible/ansible caught my eye on GitHub Trending today. 🔗 https://github.com/ansible/ansible 📝 Ansible is a radically simple IT automation platform that makes your applications and systems easier to deploy and maintain. Automate everything from code deployment to network configuration to cloud management, in a language that approaches plain English, using SSH, with no agents to install on remote systems.https://docs.ansible.com. ────────────────────────────── Ansible is a simple IT automation system that handles configuration management, application deployment, and more. Its key features include an extremely simple setup process, agentless architecture, and the ability to describe infrastructure in a human-friendly language. To use Ansible, you can install it with pip or a package manager, and power users can run the devel branch for the latest features. The community is active, with a forum for asking questions, getting help, and interacting with other users. From a technical perspective, Ansible focuses on security and auditability, and allows module development in any dynamic language. The project is coded in a variety of languages, including Python, and has a devel branch for ongoing development. Ansible is suitable for a wide range of users, from system administrators to developers, and is widely used in the industry. The project is licensed under the GNU General Public License v3.0 or later. In short, Ansible is all about making complex IT tasks radically simple - and that's a game-changer! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🎯 iv-org/invidious landed on trending. Worth a proper look. 🔗 https://github.com/iv-org/invidious 📝 Invidious is an alternative front-end to YouTube ────────────────────────────── Invidious is an open-source alternative front-end to YouTube, offering a lightweight and ad-free experience. Its key features include no tracking, customizable homepage, and audio-only mode. Users can import subscriptions from YouTube and other platforms, and export them as needed. From a technical standpoint, Invidious has an embedded video support and a developer API, making it a great option for developers. The project is hosted on GitHub and has a large community of contributors, with translations available in many languages. To get started, users can select a public instance or host Invidious themselves by following the installation instructions. The project is suitable for anyone looking for a private and customizable YouTube experience. Invidious is perfect for those who want to ditch YouTube's ads and tracking - and it's completely free and open-source. Join the Invidious community today and experience the power of open-source video sharing! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

⚡ abus-aikorea/voice-pro is making waves. Here's the full picture. 🔗 https://github.com/abus-aikorea/voice-pro 📝 Gradio WebUI for creators and developers, featuring key TTS (Edge-TTS, kokoro) and zero-shot Voice Cloning (E2 & F5-TTS, CosyVoice), with Whisper audio processing, YouTube download, Demucs vocal isolation, and multilingual translation. ────────────────────────────── Voice-Pro is an AI-powered web application designed for speech recognition, translation, and multilingual dubbing. Its key features include top-tier speech recognition with Whisper, Faster-Whisper, and Whisper-Timestamped, as well as zero-shot voice cloning with F5-TTS, E2-TTS, and CosyVoice. The application also supports multilingual text-to-speech with Edge-TTS and kokoro, and offers instant translation for over 100 languages. To use Voice-Pro, simply download and install the application, then access its web interface to upload your media files and select the desired features. The application is designed to be user-friendly and accessible, with a clean and intuitive interface. From a technical standpoint, Voice-Pro is built using a range of cutting-edge technologies, including Python 3.12, Torch 2.8.0+cu128, and Gradio 6.20. It also utilizes several AI models, including Whisper, F5-TTS, and CosyVoice, to deliver high-quality speech recognition and voice cloning capabilities. Voice-Pro is designed for a range of users, including creators, researchers, and multilingual professionals. It offers a robust alternative to ElevenLabs, with advanced voice solutions and a user-friendly interface. In short, Voice-Pro is a powerful tool for anyone looking to leverage the power of AI for speech recognition, translation, and multilingual dubbing - and with its open-source code and free distribution, the possibilities are endless! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe