Github Top Repositories

🔥 PostHog/posthog is trending — and it deserves your attention. 🔗 https://github.com/PostHog/posthog 📝 🦔 PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP. ────────────────────────────── PostHog is an open-source platform for building self-driving products, providing a range of tools to help you understand user behavior and optimize your product. With PostHog, you can capture product analytics, web analytics, and session replays, as well as set up feature flags, experiments, and error tracking. The platform also includes a data warehouse, data pipelines, and AI observability features. To get started with PostHog, you can sign up for a free account on PostHog Cloud or self-host the platform using Docker. The platform offers a generous free tier and transparent pricing for its paid plan. PostHog has a large community of contributors and offers extensive documentation, including a company handbook, product guides, and developer resources. The platform is written in a variety of languages, including JavaScript, Python, and Node, and has SDKs and libraries for popular frameworks like React and Angular. Whether you're a product manager, developer, or data analyst, PostHog has the tools you need to build a successful product. So why not give it a try and see how PostHog can help you drive growth and revenue? With its powerful features and flexible pricing, PostHog is the perfect choice for anyone looking to take their product to the next level: build better products, faster. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🎯 github/copilot-sdk landed on trending. Worth a proper look. 🔗 https://github.com/github/copilot-sdk 📝 Multi-platform SDK for integrating GitHub Copilot Agent into apps and services ────────────────────────────── The GitHub Copilot SDK allows you to embed Copilot's agentic workflows into your application. It exposes the same engine as the Copilot CLI, handling planning, tool invocation, and file edits. The SDK is available for Python, TypeScript, Go, .NET, Java, and Rust, with each implementation providing a production-tested agent runtime that can be invoked programmatically. To get started, you can install the SDK using the provided commands, then define agent behavior and let Copilot handle the rest. The SDK communicates with the Copilot CLI server via JSON-RPC, managing the CLI process lifecycle automatically. Key features include support for BYOK (Bring Your Own Key), custom agents, skills, and tools, as well as multiple authentication methods. The SDK is production-ready and follows semantic versioning. Whether you're a developer looking to streamline your workflow or a business seeking to automate tasks, the GitHub Copilot SDK is a powerful tool to consider. With its robust features and ease of use, it's an ideal solution for anyone looking to harness the power of Copilot in their application. The GitHub Copilot SDK is a game-changer for developers: it puts the power of AI directly into your code. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🌟 andrewrabert/jellium-desktop caught my eye on GitHub Trending today. 🔗 https://github.com/andrewrabert/jellium-desktop 📝 An unofficial desktop client for Jellyfin ────────────────────────────── Jellium Desktop is an unofficial desktop client for Jellyfin, built on top of CEF and mpv. It provides an easy-to-use interface for managing and streaming media content. The app is available for download on Linux, macOS, and Windows platforms, with various installation options such as AppImage, Flatpak, and Arch Linux (AUR). To get started, users can download the app from the provided links and follow the installation instructions. On macOS, users need to remove the quarantine flag using the command sudo xattr -cr /Applications/Jellium\ Desktop.app. From a development perspective, the project utilizes just as a command runner, providing various recipes for building, testing, and maintaining the app. Some of the available recipes include:

just build
just run
just test
just fmt

The app is designed for Jellyfin users looking for a seamless desktop experience. Jellium Desktop: stream your media, simplified. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🔥 KnockOutEZ/wigolo is trending — and it deserves your attention. 🔗 https://github.com/KnockOutEZ/wigolo 📝 The go-to web for your AI coding agent — local-first search, fetch, crawl & research over MCP. No API keys, no cloud, $0/query. Public beta. ────────────────────────────── wigolo is a local-first web intelligence tool for AI agents, offering a range of features like search, fetch, crawl, extract, cache, and find-similar. It's designed to work with various agents, including Claude Code, Cursor, and Codex, and can be used as an MCP server, REST endpoint, or embedded through an SDK. Key features include: - search: multi-engine web search with rank fusion and ML reranking - fetch: load URLs through a tiered router with auto-escalation to a headless browser engine - crawl: multi-page crawl with per-domain rate limits and robots.txt respect - extract: structured data extraction from pages, including tables, metadata, and JSON-LD Technical highlights: - Runs on Node ≥ 20 with ~1.5 GB of free disk space - Supports various platforms, including macOS, Linux, and Windows - npx wigolo init command sets up the local engine, downloading the browser engine and on-device models - npx wigolo doctor checks the health of the setup Audience: - AI agent developers - Users of Claude Code, Cursor, Codex, and other supported agents - Anyone looking for a local-first web intelligence solution Takeaway: With wigolo, you can empower your AI agents with a robust, local-first web intelligence layer, free from API keys, cloud dependencies, and metered bills. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🚀 Meet jamiepine/voicebox: a gem from today's GitHub trending list. 🔗 https://github.com/jamiepine/voicebox 📝 The open-source AI voice studio. Clone, dictate, create. ────────────────────────────── The Voicebox is an open-source AI voice studio that allows you to clone any voice, generate speech in 23 languages, dictate into any app, and give AI agents a voice of your choice. It's a local-first alternative to other cloud-based services, offering complete privacy as all models, voice data, and captures never leave your machine. Key features include 7 TTS engines, voice cloning and preset voices, post-processing effects, and unlimited generation length. The app also supports global dictation with a hotkey, voice input, and agent voice output. From a technical standpoint, Voicebox is built with Tauri (Rust) and has a REST API plus a built-in MCP server for integration into other apps and agents. The target audience for Voicebox appears to be developers, content creators, and anyone interested in AI-powered voice technology. In a nutshell, Voicebox is a powerful tool that puts you in control of your voice data and offers a wide range of features for voice cloning, generation, and input - and it's all free and open-source. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

📌 Spotted on GitHub Trending: rohitg00/ai-engineering-from-scratch — let's break it down. 🔗 https://github.com/rohitg00/ai-engineering-from-scratch 📝 Learn it. Build it. Ship it for others. ────────────────────────────── The AI Engineering from Scratch GitHub repository is a comprehensive curriculum designed to teach AI engineering skills from the ground up. It consists of 20 phases and 503 lessons, covering a wide range of topics from math foundations to autonomous systems. The curriculum is structured to provide a linear learning experience, with each lesson building on the previous one, and includes Python, TypeScript, Rust, and Julia code examples. The key features of this curriculum include: * A focus on building AI systems from scratch, rather than just using pre-built libraries and frameworks * A comprehensive coverage of AI topics, including math foundations, machine learning, deep learning, and autonomous systems * A linear learning structure, with each lesson building on the previous one * A variety of code examples in different programming languages To get started with the curriculum, you can choose from three options: * Read the lessons online * Clone the repository and run the code examples * Use the built-in agent skills to find your level and get personalized recommendations The curriculum is designed for anyone who wants to learn AI engineering, from beginners to experienced practitioners. The prerequisites are minimal, requiring only basic programming skills and a willingness to learn. The technical highlights of the curriculum include: * A focus on building reusable tools and artifacts, rather than just completing exercises * A comprehensive coverage of AI topics, including math foundations, machine learning, and deep learning * A variety of code examples in different programming languages Overall, the AI Engineering from Scratch curriculum is a valuable resource for anyone who wants to learn AI engineering skills from the ground up. With its comprehensive coverage of AI topics, linear learning structure, and focus on building reusable tools and artifacts, it provides a unique and valuable learning experience. Takeaway: With AI Engineering from Scratch, you don't just learn AI - you build it, from scratch, and ship reusable tools and artifacts that you can use in your daily workflow. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

⚡ kvcache-ai/ktransformers is making waves. Here's the full picture. 🔗 https://github.com/kvcache-ai/ktransformers 📝 A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations ────────────────────────────── The KTransformers project is a flexible framework for experiencing cutting-edge large language model (LLM) inference and fine-tuning optimizations. It focuses on efficient inference and fine-tuning through CPU-GPU heterogeneous computing. The project exposes two key capabilities: inference and fine-tuning with LLaMA-Factory integration. Inference is powered by the high-performance kt-kernel, which offers CPU-optimized kernel operations for heterogeneous LLM inference. This includes features like AMX/AVX acceleration, MoE optimization, and quantization support. For fine-tuning, KTransformers integrates with LLaMA-Factory, allowing for ultra-large MoE model fine-tuning. This integration supports multi-backend, ultra-large MoE models, and faster training speeds. The project is developed and maintained by several teams, including the MADSys Lab at Tsinghua University and community contributors. KTransformers welcomes contributions and provides support through GitHub issues and a WeChat group. To get started, users can follow the quick-start guides for both inference and fine-tuning, which include installing the necessary packages and launching the training process. KTransformers is ideal for researchers and developers working with large language models, particularly those interested in efficient inference and fine-tuning on various hardware configurations. In summary, KTransformers is a powerful tool for optimizing LLM inference and fine-tuning - experience the future of AI with KTransformers. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🌟 tirth8205/code-review-graph caught my eye on GitHub Trending today. 🔗 https://github.com/tirth8205/code-review-graph 📝 Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows. ────────────────────────────── The code-review-graph GitHub repository offers a solution to optimize code review tasks for AI coding tools. It creates a structural map of your codebase using Tree-sitter, tracks changes, and provides context to AI assistants via the Model Context Protocol (MCP), reducing the amount of code they need to read. The tool supports various platforms, including Codex, Claude Code, and GitHub Copilot, and can be installed with a single command. The

code-review-graph install

command auto-detects and configures supported platforms, and the tool can be used with Python 3.10+. It also features incremental updates, blast-radius analysis, and broad language coverage, including support for Jupyter notebooks. Developers can add their own languages without forking the repository by creating a languages.toml file. The repository also includes a GitHub Action for risk-scored PR reviews in CI. The tool has been benchmarked across six real repositories, with a median per-question token reduction of ~82x. The code-review-graph is a valuable resource for developers looking to optimize their code review workflow. One-liner takeaway: code-review-graph revolutionizes code reviews by giving AI assistants precise context, slashing token waste, and making the development process smarter and more efficient. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🎯 MoonshotAI/kimi-cli landed on trending. Worth a proper look. 🔗 https://github.com/MoonshotAI/kimi-cli 📝 Kimi Code CLI is your next CLI agent. ────────────────────────────── Kimi CLI is an AI agent that runs in the terminal, helping you complete software development tasks and terminal operations. It can read and edit code, execute shell commands, search and fetch web pages, and autonomously plan and adjust actions during execution. Key features include a shell command mode, VS Code extension, IDE integration via ACP, Zsh integration, and MCP support. You can use Kimi CLI with various tools and editors, such as Visual Studio Code, Zed, and JetBrains, to enhance your development experience. To get started, you can install Kimi CLI and refer to the getting started guide. For development, you can clone the repository, prepare the environment, and run Kimi CLI using uv run kimi. Technical highlights include support for MCP tools, ad-hoc MCP configuration, and a range of sub-commands for managing MCP servers. The project is evolving into Kimi Code CLI, which will automatically migrate your configuration and sessions. This project is suitable for developers and users who want to leverage AI capabilities in their terminal and development workflow. One-liner takeaway: Kimi CLI is an AI-powered terminal agent that streamlines your development workflow, and it's evolving into Kimi Code CLI for even more capabilities. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

🚀 Meet codecrafters-io/build-your-own-x: a gem from today's GitHub trending list. 🔗 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 step-by-step guides for building a wide range of technologies from scratch. With the motto "What I cannot create, I do not understand" from Richard Feynman, this project encourages learning through hands-on creation. The repository offers tutorials and guides on building various technologies, including 3D renderers, AI models, augmented reality, blockchains, bots, command-line tools, databases, and many more. These guides cover a range of programming languages, such as C, C++, Java, Python, and JavaScript, making it accessible to developers with different skill sets. The guides are well-structured and provide a comprehensive overview of the technology being built, including the technical aspects and implementation details. They also offer a unique opportunity for developers to learn by doing and gain a deeper understanding of the technologies they use. The target audience for this repository appears to be developers and programmers who want to learn and improve their skills by building real-world projects from scratch. Whether you're a beginner or an experienced developer, this repository has something to offer. In short, the codecrafters-io/build-your-own-x repository is an excellent resource for anyone looking to learn and build new skills, and its motto says it all: build it to understand it. Don't just use it, build it! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

💡 KnockOutEZ/wigolo just hit the trending charts — here's why it matters. 🔗 https://github.com/KnockOutEZ/wigolo 📝 The go-to web for your AI coding agent — local-first search, fetch, crawl & research over MCP. No API keys, no cloud, $0/query. Public beta. ────────────────────────────── Meet wigolo, a local-first web intelligence platform designed for AI agents. It provides a durable surface for everything web-related, including search, fetch, crawl, extract, cache, and research, all without requiring API keys or incurring metered bills. Key features include multi-engine web search, tiered routing for fetching pages, structured data extraction, and a memory that compounds with each query. wigolo runs wherever your agent runs, whether as an MCP server, a REST/MCP endpoint, or embedded through an SDK. To get started, simply run npx wigolo init --agents=<your-agent> to set up the local engine. You can then use various tools like search, fetch, and research to gather information. Technical highlights include on-device models, direct adapters for public engines, and transparent per-result scoring. wigolo is designed for agents, not humans, and provides honest output with surfaced degradation and self-flagged junk results. Audience: This platform is ideal for developers and users of AI agents, including those using Claude Code, Cursor, Codex, and other popular AI tools. In short, wigolo is a powerful, private, and free web intelligence platform that empowers your AI agents to gather information without incurring costs or relying on third-party services. The takeaway: with wigolo, your AI agents can search, fetch, and research the web without breaking the bank or sacrificing privacy. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

💡 lyogavin/airllm just hit the trending charts — here's why it matters. 🔗 https://github.com/lyogavin/airllm 📝 AirLLM 70B inference with single 4GB GPU ────────────────────────────── AirLLM is a game-changer for large language models, enabling you to run massive models like 405B Llama 3.1 on a single 8GB GPU and 671B DeepSeek-V3 on ~12GB. This is achieved without quantization, distillation, or pruning. The key to this magic is that AirLLM only keeps one layer on the GPU at a time, reducing the required VRAM significantly. To get started, you can install the airllm package using pip install airllm. Then, initialize the model using AutoModel.from_pretrained(), passing in the Hugging Face repo ID or local path of the model. You can also enable model compression for up to 3x inference speedup by specifying the compression argument. AirLLM supports a wide range of models, including Llama, Qwen, DeepSeek, Mistral, and many more. It's perfect for developers and researchers who want to work with large language models without breaking the bank on GPU hardware. In short, AirLLM is a powerful tool that makes large language models accessible to everyone. With its ease of use and impressive performance, you can now run huge models on relatively small GPUs - unlocking new possibilities in NLP research and development. ────────────────────────────── 🧠 Channel: https://t.me/GithubRe

💡 elder-plinius/G0DM0D3 just hit the trending charts — here's why it matters. 🔗 https://github.com/elder-plinius/G0DM0D3 📝 LIBERATED AI CHAT ────────────────────────────── The G0DM0D3 project is a fully open-source, privacy-transparent, multi-model chat interface that pushes the limits of post-training layer capabilities. It's designed for hackers, philosophers, and system tinkerers, offering features like multi-provider support with 60 listed OpenRouter models, GODMODE CLASSIC for racing model combos in parallel, and ULTRAPLINIAN for multi-model evaluation. Key features include: - Parseltongue for input perturbation and red-teaming research - AutoTune for context-adaptive sampling parameters - Local History for conversation storage and export/import support - Responsive design for desktop and mobile use To get started, visit the hosted site at godmod3.ai or self-host the standalone interface by cloning the repository and opening index.html in your browser. For local model support, configure OpenRouter, Venice, or your own models in Settings. The project emphasizes privacy controls, with metadata-only app telemetry that can be disabled with No-Log or Local-only mode. Chat history is stored in browser storage, and there's no account system or cloud history sync. G0DM0D3 is perfect for those who want to explore the frontiers of AI interaction without sacrificing control or privacy. The punchy one-liner takeaway: Experience liberated AI cognition with G0DM0D3, where you're in control, not the AI overlords! ────────────────────────────── 🧠 Channel: https://t.me/GithubRe