📦 averygan/reclip

Stop struggling to save your favorite videos online

Tired of dealing with broken download buttons and annoying ads whenever you want to save a video? You need to check out Reclip, a powerful self-hosted tool that lets you download media from almost any website with total ease. It gives you a clean web interface to manage your downloads directly, keeping everything organized and lightweight without any clutter. By running it on your own hardware, you gain complete control over your media collection and privacy. Whether you are archiving content or grabbing clips for later, Reclip simplifies the entire process. Try it out and take back your library today.

📰 https://news.ycombinator.com/item?id=48049953

🆔 @hackernewsgithubprojects

📦 jarczakpawel/orcaslicer-bambulab

The Truth Behind the OrcaSlicer Bambu Lab Conflict

Are you curious why a major 3D printing project suddenly vanished? The repository for the OrcaSlicer integration with Bambu Lab hardware became a focal point of intense legal pressure and corporate pushback. The developer faced direct demands to take down their work, which aimed to bridge the gap between proprietary Bambu Lab systems and open-source workflows. After being accused of bypassing authorization controls and reverse engineering, the creator chose to remove the software entirely to avoid a prolonged legal battle. It serves as a fascinating, if cautionary, lesson on the ongoing tension between open-source enthusiasts and restrictive manufacturer ecosystems.

📰 https://news.ycombinator.com/item?id=47878153

🆔 @hackernewsgithubprojects

📦 dyoshikawa/rulesync

Stop Managing AI Prompts Manually with This Tool

Are you tired of manually updating configuration files every time you switch between different AI coding agents? Keeping your instructions and rule files consistent across tools like Cursor, Claude Code, and Copilot can be a total headache. That is where a smart utility called Rulesync comes in. It acts as a single source of truth, letting you define your rules once and automatically generating the specific config files for every AI tool in your stack. Whether you need to import existing setups or convert rules between different agents, this project keeps your development environment perfectly synced and ready to go.

📰 https://news.ycombinator.com/item?id=48051242

🆔 @hackernewsgithubprojects

📦 antirez/ds4

Run DeepSeek V4 Flash Locally on Your Mac

Want to run a powerful AI model directly on your Apple hardware? This repository provides a dedicated, lightweight inference engine specifically designed for DeepSeek V4 Flash using Metal. Unlike generic frameworks, it is built for extreme efficiency, featuring advanced capabilities like on-disk KV cache persistence, which allows you to keep context across sessions without eating up all your RAM. It is a fantastic tool for developers wanting to run large, complex models on high-end personal workstations. Check out this specialized engine to experience blazing-fast, local AI performance that treats your SSD like first-class memory for long-context tasks.

📰 https://news.ycombinator.com/item?id=48050751

🆔 @hackernewsgithubprojects

📦 vectorize-io/hindsight

Give Your AI Agents a Real Memory with Hindsight

Ever wish your AI agents could actually learn from their past experiences instead of just resetting every time? Meet Hindsight, a powerful open source memory system designed to help AI agents evolve. Unlike basic systems that only store conversation history, Hindsight uses biomimetic structures to categorize information into facts, experiences, and mental models. By leveraging tools like retain, recall, and reflect, your agents can actively analyze data and build deeper insights over time. Whether you use the simple SDK or the LLM wrapper, Hindsight makes upgrading your agent with long term, actionable memory remarkably easy. Start building smarter, more capable agents today.

🆔 @hackernewsgithubprojects

📦 pthom/imgui_bundle

Stop Fighting GUI Frameworks with Dear ImGui Bundle

Are you tired of battling complex graphical user interface frameworks just to build a simple tool? Meet the perfect solution for creating interactive applications. This powerhouse library bundles Dear ImGui with a collection of essential tools for both C plus plus and Python developers. It provides seamless integration for 2D and 3D plotting, advanced image inspection, and even node editors, all pre-wired to work immediately. Whether you are targeting desktop, mobile, or the web, this project lets you focus on building features instead of managing dependencies. Stop struggling with interface code and start bringing your creative ideas to life today.

📰 https://news.ycombinator.com/item?id=48052290

🆔 @hackernewsgithubprojects

📦 janpfeifer/gonb

Code Go in Jupyter Notebooks with GoNB

If you love the Go programming language but wish you could use it inside Jupyter notebooks, you need to check out GoNB. It is a powerful interactive kernel that lets you write, run, and experiment with Go code right in your browser just like you do with Python. By providing seamless integration with standard notebook features, it solves the headache of constant compiling when you just want to test snippets or visualize data. You can even use it to create rich HTML output or plot graphs directly from your code. It is the perfect tool for making your Go development interactive.

📰 https://news.ycombinator.com/item?id=48052434

🆔 @hackernewsgithubprojects

📦 iandchasse/de-link

Build Your Own Custom E-Reader with the de-link Kit

Ever wanted to build your own custom e-reader? Meet the de-link, an open-source development kit powered by the ESP32-S3 that puts you in full control of your hardware. It solves the frustration of locked-down proprietary devices by providing a customizable platform for 24-pin e-paper displays. You get features like 4-bit SDMMC for storage, USB-C connectivity, and modular options for cool and warm frontlight control. It is designed to be hand-soldered, making it perfect for makers who want to repair or upgrade their own gear. Grab your soldering iron and start creating a truly personal reading experience today.

📰 https://news.ycombinator.com/item?id=48052657

🆔 @hackernewsgithubprojects

📦 scosman/cursed_browser

The Web Browser That Doesn't Actually Render Anything

Imagine trying to navigate the internet without a screen or a real browser engine to render the pages. This bizarre project takes a radical approach by feeding raw HTML code directly into a vision language model. Instead of standard rendering, the AI analyzes the structure and creates a hallucinated version of the webpage for you to interact with. It is a fascinating experiment in artificial intelligence that challenges how we perceive digital information. While it might not replace your daily driver, it is a brilliant look at how machines interpret the web. Try building one yourself for a truly wild experience.

📰 https://news.ycombinator.com/item?id=48010670

🆔 @hackernewsgithubprojects

📦 zhangfengcdt/memoir

Give Your AI Agent Git-Like Memory with Memoir

Stop letting your AI agent lose track of critical details with messy context storage. Meet Memoir, a powerful tool that brings Git-like version control to your agent memory. Instead of using flat, confusing files, Memoir uses hierarchical semantic paths to organize information with cryptographic integrity. You can branch, commit, merge, and rollback memories just like code, which prevents context contamination and hallucinations during complex tasks. Whether you are managing professional preferences or coding styles, this system provides efficient log-time lookups and a clean visual explorer to keep your AI smart and reliable. Upgrade your agent workflow today and master your memory.

📰 https://news.ycombinator.com/item?id=48024243

🆔 @hackernewsgithubprojects

📦 mvanhorn/cli-printing-press

Build the Ultimate AI-Powered CLI in Minutes

Stop wasting time and tokens on clunky API interactions because this project is a game changer for AI agent workflows. It automatically discovers your API's hidden potential, absorbing features from every competing tool to build the ultimate CLI. Designed specifically for AI agents, it integrates local SQLite databases for instant offline searching and compound insight commands that standard wrappers simply cannot handle. Whether you need a native Go CLI or an MCP server, this tool generates both from a single specification, giving you incredible speed and efficiency. Take control of your developer workflow and start printing better CLIs today.

📰 https://news.ycombinator.com/item?id=48054795

🆔 @hackernewsgithubprojects

📦 manojmallick/sigmap

Stop Wasting AI Context: Meet SigMap

Is your AI agent constantly guessing because it lacks codebase awareness? Stop burning through your token budget on redundant file reads. SigMap is a powerful, zero-dependency tool that generates a compact signature map of your entire project, cutting down context usage by up to ninety-seven percent. By extracting only the essential function and class signatures across twenty-nine different languages, it ensures your AI tools like Cursor or Copilot always focus on the right files. It even includes a dedicated MCP server for intelligent retrieval, keeping your development workflow lightning fast and incredibly accurate. Upgrade your coding experience and start building smarter today.

📰 https://news.ycombinator.com/item?id=47956790

🆔 @hackernewsgithubprojects

📦 nvlabs/cuda-oxide

Write GPU Kernels in Pure Rust with Cuda-Oxide

Are you tired of writing complex CUDA code that feels disconnected from your main project? Meet the experimental compiler that is changing the game by letting you write GPU kernels in pure, idiomatic Rust. Instead of dealing with foreign language bindings or messy domain-specific languages, this tool allows your host and device code to live together in the same file. It compiles directly to machine-ready code, supporting advanced features like shared memory, atomics, and even closures with captures. If you want to harness the power of your GPU while keeping the safety and expressiveness of Rust, give this project a look.

📰 https://news.ycombinator.com/item?id=48055332

🆔 @hackernewsgithubprojects

📦 codecrafters-io/build-your-own-x

Master Programming by Building Projects From Scratch

Ever wonder how your favorite software actually works under the hood? Instead of just using tools, you can master programming by recreating them from scratch. This collection of guides helps you peel back the layers of technology, whether you want to build your own database from the ground up, develop a functional blockchain, or even write a custom operating system. By tackling these complex challenges step by step, you bridge the gap between theory and practice, gaining a deeper understanding of computer science concepts. Pick a project today and start building your way to becoming a better developer.

📰 https://news.ycombinator.com/item?id=48057164

🆔 @hackernewsgithubprojects

📦 getbindu/awesome-claude-code-and-skills

Supercharge Your Coding with Claude Skills

Are you tired of manually handling repetitive coding tasks? If you use Claude for development, you need to see this incredible curated collection of Claude Skills. This repository acts as a massive toolkit that helps you extend the functionality of your AI assistant through specialized agents, plugins, and custom configurations. Whether you need automated code reviews, advanced security auditing, or even full multi-agent orchestration, these production-ready skills provide the structure to transform your workflow. By implementing these pre-built modules, you save countless hours and gain professional-grade automation. Start exploring these powerful community resources today to build faster and smarter.

📰 https://news.ycombinator.com/item?id=47562741

🆔 @hackernewsgithubprojects

📦 pacifaist/quansloth

Run Massive AI Models on Any GPU with Quansloth

Stop letting your GPU crash when running massive AI models. Quansloth is a private, air-gapped AI server that uses elite TurboQuant KV cache compression to break through the memory wall on consumer hardware. By shrinking your model's memory footprint by up to seventy-five percent, it allows you to process massive documents on modest GPUs that would typically crash. It bridges a sleek Gradio frontend with an optimized CUDA backend for stable, long-form analysis. Whether you are a privacy enthusiast or a power user, Quansloth makes high-end AI inference accessible and stable. Take back control of your local AI workflow today.

📰 https://news.ycombinator.com/item?id=47672422

🆔 @hackernewsgithubprojects

📦 darkrishabh/agent-skills-eval

Master Your AI Agent Testing

Are you struggling to verify if your AI agents truly understand their specific tasks? This project provides a robust test runner designed to evaluate agent skills with precision. By standardizing how you validate automated behaviors, it solves the major pain point of inconsistent performance in complex AI workflows. It allows you to run structured tests that verify capability outcomes against expected benchmarks, ensuring your agents remain reliable and efficient. Whether you are building advanced digital assistants or custom automation tools, this evaluation framework is an essential resource for quality control. Start testing your agents today to achieve more consistent and predictable results.

📰 https://news.ycombinator.com/item?id=48046023

🆔 @hackernewsgithubprojects

📦 quoll/asami

Asami: The Flexible Graph Database for Clojure

Ever struggle with rigid database schemas that slow down your development process? Meet Asami, a powerful and flexible graph store written for Clojure and ClojureScript. Asami solves the problem of rigid data modeling by allowing you to store information without predefined schemas, making it perfect for evolving projects where data structures change over time. Its immutable storage ensures stability, while its query API feels intuitive and familiar to those who have used systems like Datomic. Whether you are running in the browser or on the server, Asami provides a seamless way to manage interconnected data. Dive into graph-based development today.

📰 https://news.ycombinator.com/item?id=48060970

🆔 @hackernewsgithubprojects

📦 cognitect-labs/anomalies

Stop Using Exceptions: A Better Way to Handle Errors

Are your error handling routines feeling bloated and messy? Meet the anomalies library for Clojure, a game-changer for managing errors without relying on clunky stack unwinding. Instead of throwing exceptions that disrupt your control flow, this project treats errors as simple, actionable maps. By using standardized categories, you can easily determine if a failure is retryable, like when a service is busy, or if it requires a fix to your caller logic. It makes your code cleaner, more predictable, and easier to debug. Start treating errors as data today and watch your system stability reach entirely new levels.

📰 https://news.ycombinator.com/item?id=48060942

🆔 @hackernewsgithubprojects

📦 wojtczyk/trust

Coding Rust Like It's 1989

Have you ever wondered what the modern power of Rust would look like if it were designed in the late eighties? This fascinating project experiments with a unique aesthetic by stripping away contemporary conventions to embrace a raw, retro development style. By prioritizing simple text interfaces and minimalist tooling, it creates an unconventional workflow that feels like a classic terminal experience while still leveraging the safety and performance of a modern language. It is a brilliant way to challenge your assumptions about developer environments and explore how constraints can reshape your code. Give it a look to rethink your stack.

📰 https://news.ycombinator.com/item?id=48045939

🆔 @hackernewsgithubprojects

📦 omerbenamram/mft

Master File Table Parsing Made Easy

Ever wondered how to unlock the secrets hidden deep within your hard drive? The MFT project is a powerful, high-performance tool designed to parse the Master File Table, which acts as the roadmap for your entire NTFS filesystem. By converting complex MFT snapshots into clean JSON or CSV formats, it makes forensic investigation and system analysis incredibly efficient. Written entirely in safe Rust, this parser handles large volumes of data with impressive speed and precision. Whether you need to extract resident data streams or perform deep metadata analysis, this tool provides exactly the clarity you need to master your data.

📰 https://news.ycombinator.com/item?id=48062233

🆔 @hackernewsgithubprojects