📦 trailhq/graft

Graft: Give AI a Memory Map

Graft turns your entire codebase into a readable map that AI agents actually understand, cutting their work time by sixty percent. Usually, coding tools start every new task by blindly searching through files, wasting time and money. This tool solves that by building a folder of linked notes that explain exactly what every part of your system does in plain English. It is not a hidden database, but a local cache that stays fresh automatically. Since it uses standard files, any tool can read it instantly. You get faster results and better answers without changing your setup.

🆔 @hackernewsgithubprojects

📦 zorost/ai-engineering-lab

AI Engineering Lab

AI Engineering Lab is the self-paced course that treats you like a real engineer from day one. Instead of jumping between random tutorials, you build one continuous project for a fictional freight company over twenty-four weeks. That single story keeps your code connected, so your early data becomes the foundation for your final cloud deployment. It is free, fully open, and requires no signup. If you want to stop guessing what to learn next and start building systems that actually work, this is the map you need.

🆔 @hackernewsgithubprojects

📦 useagenthq/useagent

useagent: The AI Coworker

Your AI agent just got a full computer and a job description. useagent is an open source platform that turns coding assistants into actual teammates who can build real things like websites, spreadsheets, and code changes. Instead of just chatting, these agents work inside secure cloud machines, using your specific tools and company knowledge to finish tasks from start to finish. You can even run them right from Slack, just like pinging a human colleague. The coolest part is that every action is logged, so you can always see exactly what the agent did and why. It’s basically hiring a tireless intern who never sleeps and never forgets a step.

🆔 @hackernewsgithubprojects

📦 kevinliss/apeadmin

ApeAdmin: The Admin Panel That Speaks AI

ApeAdmin is an open-source Python admin framework that lets you control your entire application through AI. It wraps standard features like user management and permission settings into tools that language models can call directly, meaning your AI assistant can perform real actions inside your system, not just chat about them. It includes an easy installation wizard, a plugin marketplace for adding new features, and built-in security to ensure the AI only does what you allow. If you are building apps where people and machines work together, this is the bridge that connects them. It is a powerful, practical way to build modern software today.

🆔 @hackernewsgithubprojects

📦 trailofbits/coop

Coop: The VM for Coding Agents

Coop is the isolated VM manager that keeps your coding agents safely fenced in. It spins up disposable virtual machines where tools like Claude Code run with full access to compilers and Docker, but without ever touching your actual computer. The real surprise is how it handles your secret API keys. Instead of letting the agent see them, Coop injects credentials through a secure proxy that sits outside the machine. This means if an agent gets confused or malicious, it simply cannot steal your passwords. It is the perfect sandbox for letting AI code freely while keeping your host machine completely untouched.

📰 https://news.ycombinator.com/item?id=49593842

🆔 @hackernewsgithubprojects

📦 tight-line/ballast

Ballast: The Kubernetes Operator That Actually Fixes Your Resource Wastage

Stop guessing how much your apps need and let Ballast do the math for you. This Kubernetes operator watches your workloads, learns exactly how much they use, and then actually changes their settings to match reality. Most tools just guess or give you advice, but Ballast applies those fixes directly, even to running containers. The coolest part is that it remembers everything across your whole cluster. If you deploy the same app in forty different places, they all share one smart history. That means your new deployments start perfectly sized from the very first second, not from a cold start.

📰 https://news.ycombinator.com/item?id=48767949

🆔 @hackernewsgithubprojects

📦 fif7y/pelmet

Pelmet: Clean Up Your Mac Menu Bar

Tuck away your cluttered menu bar and reclaim your desktop space. Pelmet is a lightweight tool for Macs that hides the icons you rarely use until you actually need them. It works directly with the system, so sliding icons in and out feels completely natural, with no lag or weird visual glitches. You decide exactly what stays visible and what tucks away behind a simple chevron. It’s free, open source, and runs entirely on your computer without sending data anywhere. If your menu bar feels crowded and distracting, this is the calm solution you have been looking for.

🆔 @hackernewsgithubprojects

📦 fredyrivera-dev/claude_converter

Claude Converter: Turn AI Chats into Training Data

Claude Converter takes the messy chat logs from your coding AI sessions and turns them into clean data that standard language models can actually read. You know those long files where your AI assistant solves problems? This tool grabs those records, strips away the confusing formatting, and flattens them into a simple list of questions and answers. It works for three different AI coding tools, so you don't have to learn a new script for each one. This is perfect if you want to use your own past conversations to teach a smaller model how to code.

🆔 @hackernewsgithubprojects

📦 nicologiuliani6/cavium-cn6640-snic10e-octeon-ii-nic

Cavium CN6640-SNIC10E Octeon II NIC: A Working Linux Driver

Turn a stubborn industrial network card into two full speed gigabit ethernet ports on your Linux machine. This project builds a custom driver that bypasses the manufacturer’s broken firmware, allowing a computer to directly control the card’s hardware through a shared memory trick. It was previously impossible to use this specific card as a standard network interface because its internal registers were locked and unreadable to the host. This repository solves that by manually programming those registers from the card side, creating a direct channel for data to flow without any proprietary software.

📰 https://news.ycombinator.com/item?id=49532788

🆔 @hackernewsgithubprojects

📦 marsbos/mador

Mador: Make Any DOM Reactive

Mador is the micro-library that makes any existing HTML reactive without a framework. It solves the problem of adding interactivity to static pages without the bloat of React or Vue. You simply import a tiny script, define your state, and bind it directly to elements on the page. The most surprising part is how it handles data lists; it automatically moves and updates HTML nodes to match your data array, acting like a smart diffing engine. There is no virtual DOM and no component lifecycle to manage. It is just a small piece of code that keeps your interface in sync with your data.

📰 https://news.ycombinator.com/item?id=49590738

🆔 @hackernewsgithubprojects

📦 vibegameengine/web-starter-kit

Web Starter Kit: A Blueprint for AI-Driven 3D Game Development

Kickstart your browser-based game development with a complete, opinionated foundation for building interactive 3D web applications. This project provides a structured environment where developers can define specific tasks for artificial intelligence agents, effectively outsourcing complex workflows like character rigging, asset optimization, and scene architecture to automated specialists. It solves the problem of managing the chaotic, multi-disciplinary nature of game production by organizing AI capabilities into a coherent, collaborative team structure. By standardizing how these digital assistants interact and execute specialized skills, it turns a single developer into a coordinated studio.

🆔 @hackernewsgithubprojects

📦 optuna/rustuna

Rustuna: The Rust-Powered Speed Boost for Optuna

Rustuna is the hyperparameter optimizer that finally makes running thousands of trials feel instant. Built entirely in Rust, this project takes the popular Python tool Optuna and rewrites its core engine to run significantly faster, which is a game changer when your optimization loop itself is the bottleneck. You get the same familiar interface, so switching over is as easy as changing one import line, but the underlying speed is dramatically different. It also drops the heavy dependency on NumPy, keeping your environment lightweight and conflict-free. If you are juggling massive optimization workloads, this is the speed upgrade you have been waiting for.

🆔 @hackernewsgithubprojects

📦 swisslite/kryptos

Encrypt Your WhatsApp Messages Like a Pro

Kryptos lets you encrypt your messages right inside the chat app you already use. Instead of switching to a secure app to type, you just install it as a keyboard. When you hit the send button, the text turns into gibberish before it ever leaves your phone. This means the person receiving the message gets a blob of random characters, but Kryptos instantly decrypts it on their screen so it looks like a normal conversation. It is completely offline, so no server ever sees your data. It is the perfect way to keep your private chats truly private without changing your daily habits.

📰 https://news.ycombinator.com/item?id=49268888

🆔 @hackernewsgithubprojects

📦 gandazgul/runwield

Runwield: Keep AI Coding on Leash

Hand over the keyboard to an AI coder while keeping your fingers on the emergency brake. Runwield is a tool that lets artificial intelligence write your code changes, but it refuses to let those changes merge until you have explicitly reviewed and approved them. Instead of blindly accepting whatever the bot spits out, you get a clear, step-by-step view of its plan and the exact lines it wants to touch. This means you can watch the AI work, catch mistakes before they ship, and merge only what actually works.

🆔 @hackernewsgithubprojects

📦 kablex/comfyui-ref2va-vsa

72-Second Character Video Generation with ComfyUI Ref2VA VSA

A single consumer graphics card now generates a five-second character video in just seventy-two seconds, a speed that feels almost like a typo compared to the ten-minute standard. This ComfyUI custom node, called ComfyUI Ref2VA VSA, makes that possible by applying a clever technique called sparse attention to the MiniMax H3 model. The trick is that it aggressively ignores most of the video data to save computing power, but it carefully keeps the reference character image fully intact so the face never breaks. The result is a tool for animators who want fast, consistent character clips without renting expensive server clusters.

🆔 @hackernewsgithubprojects

📦 holymayhem/torrserve-silicon

TorrServe Silicon

TorrServe Silicon turns a raw torrent link into a proper movie card with a poster and description before you even start downloading. It runs quietly in your Mac menu bar, handling the background downloading while you just watch. No more juggling terminal commands or confusing web dashboards. You drop in a file, it figures out what movie it is, and pulls the cover art. It is a simple, native app for Apple Silicon Macs that makes torrenting look less like a hacker tool and more like your normal media library. If you use a Mac, this is the clean way to manage your downloads and start playing in one click.

🆔 @hackernewsgithubprojects

📦 chadhurley25075-png/pd-bridge

pd-bridge: How to Mix NVIDIA and Apple Silicon

pd-bridge is the inference bridge that finally lets NVIDIA and Apple Silicon work together on a single request without forcing them to speak the same language. The problem is that standard systems require both machines to agree on a memory format, but these two chips use completely different architectures. Instead of transferring a massive, unreadable cache, this system uses the fast NVIDIA box to calculate exactly what the Mac needs, then sends just that small, finished result over a standard ethernet cable. It is like hiring a specialist to do the heavy lifting and handing the finished blueprint to the local team.

🆔 @hackernewsgithubprojects

📦 ajayvarmaramineni/byoc

BYOC: Your Files, Your Rules

BYOC is the storage library that lets your app save files directly to your own cloud instead of a company server. It unifies Google Drive, Nextcloud, and S3 into one simple interface, so you never pay hidden egress fees again. The coolest part is direct browser upload: users send files straight to their personal accounts without your server ever touching the data. Your app just sees a link. This keeps your infrastructure light and gives users total control over where their documents live. If you handle user files, this is the bridge you have been missing. Stop renting storage and start using what people already own.

🆔 @hackernewsgithubprojects

📦 jonjonrankin/pi-caveman

pi-caveman: Save 75% of Tokens

Cut your AI coding costs by up to seventy-five percent without losing a single bit of technical accuracy. Meet pi-caveman, a playful extension that forces your coding agent to speak like a caveman. Instead of wasting tokens on polite filler and long explanations, it demands short, punchy fragments. It’s like hiring a brilliant engineer who only has five seconds to talk, but somehow still solves your bugs perfectly. The result is a massive drop in output costs for the same quality work, making it a clever trick for anyone who cares about efficiency.

📰 https://news.ycombinator.com/item?id=47721514

🆔 @hackernewsgithubprojects

📦 halo-box/strix-llama.cpp

strix-llama.cpp

A single chip can hold nearly one hundred and twenty eight gigabytes of shared memory, letting you run massive artificial intelligence models locally without a server room. The Strix Llama C plus project is a specialized software fork built specifically to squeeze the maximum speed out of that hardware. It rewrites the low level math routines so they fit the unique quirks of the integrated graphics processor. This means you get faster text generation and better performance right on your desk. If you have this specific hardware, this is the only tool you need to unlock its true potential.

🆔 @hackernewsgithubprojects