📦 agentconnect-md/agentconnect

AgentConnect: Team AI Agents That Talk To Each Other

AgentConnect lets multiple artificial intelligence assistants collaborate in real time inside the chat apps your team already uses. Instead of each bot working in isolation, this open source platform links them together so they can pass tasks back and forth while humans watch from a single dashboard. You connect standard coding tools like Claude or Gemini to channels like Slack and Discord, then give each assistant its own memory and set of skills. The best part is that the system runs locally on your own machines, so your private code never touches a third party server.

🆔 @hackernewsgithubprojects

📦 mewamew/my_ai_town

AI Town: A Pixel Village Driven by Artificial Intelligence

Watch a pixel village come alive where every resident actually thinks for themselves. This project builds a simulation game using Godot where characters are powered by large language models, meaning they don't just follow scripts. They wake up, go to work, chat with friends, and remember your past conversations based on their own personalities and relationships. You can fly overhead or step into the world to change the weather or send announcements, watching how your actions ripple through their social lives. It is fascinating to see how memory and context create a living, breathing community rather than a static background.

🆔 @hackernewsgithubprojects

📦 entrpi/ds4-on-spark

Run DeepSeek on a Laptop

This project lets you run the massive DeepSeek V4 model entirely on your local computer. It uses a clever trick where a small helper predicts the next few words, so the main engine only checks if it is right. This saves huge amounts of processing power. You get text about three times faster at reading your prompt, and the system handles many conversations at once without slowing down. It just needs a special NVIDIA chip and plenty of memory. It is a neat way to have a powerful AI assistant right on your desk without needing a giant server farm.

🆔 @hackernewsgithubprojects

📦 chenjin-cmd/xhs-virtual-product

xhs-virtual-product

xhs-virtual-product is the AI skill that turns Xiaohongshu virtual product creation into a repeatable workflow. It guides you through picking items, analyzing rivals, writing posts, and launching shops using ready-made templates and checklists. The real magic is the built-in compliance shield that flags copyright traps and platform rule breaks before you publish, saving you from account bans. It works with any major AI coding tool, so you just drop the files in and start building original digital goods without guessing. Grab the templates, keep it original, and let the system handle the heavy lifting.

🆔 @hackernewsgithubprojects

📦 codejunkie99/meridian-company-os

Meridian Company OS: Live Business Simulation

Meridian Company OS lets you run a fully simulated company right in your browser, where AI agents automatically send heartbeats, complete tasks, and burn through budgets every few seconds. It solves the problem of disconnected team management by giving you a single live console that tracks goals, finances, and approvals all at once. You can chat with these virtual employees to assign work, inspect their performance, and approve decisions, making it perfect for designing business workflows without needing real staff. This project is fascinating because it turns abstract operational concepts into a tangible, ticking system you can interact with immediately.

🆔 @hackernewsgithubprojects

📦 nvidia-nemo/switchyard

Switchyard: The LLM Traffic Router

Switchyard acts as a smart proxy that translates between OpenAI and Anthropic API formats while routing requests to different AI models. Built in Rust, it lets you use coding agents with open-source models by handling the messy protocol differences behind the scenes. You can split traffic for testing, track performance metrics, or write your own routing logic without touching the underlying HTTP stack. It essentially bridges the gap between your favorite developer tools and the AI backends they need to talk to.

🆔 @hackernewsgithubprojects

📦 tencent-hunyuan/hunyuan3d-buffalo1.0

Hunyuan3D-Buffalo 1.0 Explained

Tencent researchers have released a new model that turns simple text or images into fully three-dimensional objects you can rotate and edit. This is not just a static picture generator; it builds actual geometry and texture that you can tweak or modify after creation. The system handles three distinct tasks in one unified workflow, letting you generate, understand, and edit models without switching tools. It is useful because it simplifies the complex pipeline of 3D asset creation into a single step. This makes high-quality 3D content accessible for everyone. You can now create unique virtual items quickly from scratch. This is a big step forward for creative workflows.

🆔 @hackernewsgithubprojects

📦 shubhang-d/loophole

Loophole: One Tap to Fix Banking Apps

Android developers know the headache of toggling Developer Options off just to open a banking app, then remembering to turn it back on later. Loophole solves this by letting you flip that system setting with a single tap in the notification shade. It works by asking for a special permission once via a computer connection, which lets it safely write to that one specific switch. No ads, no tracking, and no network access means it just does exactly what you need without the fluff. You get a quick button in the shade and a small home screen widget to keep things simple.

📰 https://news.ycombinator.com/item?id=48974635

🆔 @hackernewsgithubprojects

📦 gtausa197-svg/-project-nord-spiking-neural-network-language-model

Project Nord: The $670 Brain-Inspired AI

Project Nord lets you run a billion-parameter language model that thinks like a brain on a simple phone. Instead of firing up every neuron like standard AI, this spiking neural network keeps ninety-three percent of its circuits perfectly silent, using only tiny bursts of spikes to process language. It is fascinating because it learned to speak from scratch without any pretrained shortcuts, proving that you can build massive intelligence for just six hundred and seventy dollars. The model even updates its memories in real time while you chat, mimicking how human neurons actually adapt. You are watching text emerge from pure biological simulation, not expensive cloud computing.

🆔 @hackernewsgithubprojects

📦 srihari7070/lore-map

Lore Map

You can now view your entire codebase as an interactive, zoomable map, edit the layout on a visual canvas, and watch Claude write the matching code automatically. Lore-map runs locally on your machine, spinning up an Express server to parse your directory structure, database tables, and import relationships. Instead of clicking through deeply nested folders, you double-click visual blocks to inspect actual file components or database schemas. You simply sketch out changes or wire up new connections on the canvas, and the tool uses the Claude Agent SDK to directly implement the edits. See your architecture, edit the map, and ship your features.

🆔 @hackernewsgithubprojects

📦 betterdb-inc/monitor

Monitor: The Missing Eye for Valkey and Redis

BetterDB Monitor saves the slow logs and command patterns that Valkey and Redis usually delete, so you can actually debug what went wrong at three in the morning instead of just staring at a blank screen. This tool acts as a historical archive for your database activity, capturing every request, latency spike, and client connection that normally vanishes after a log rotation. You get to replay live traffic, track which specific service is hammering your keys, and see exactly why the system choked an hour ago. It even spots weird behavior automatically and tells you what happened in plain English.

🆔 @hackernewsgithubprojects

📦 kyuz0/amd-strix-halo-vllm-toolboxes

amd-strix-halo-vllm-toolboxes

amd-strix-halo-vllm-toolboxes is the Fedora toolbox that finally lets you cluster multiple Strix Halo laptops into one powerful machine using fast network cards. It solves the lonely single-card problem by bundling a custom ROCm setup that connects two nodes together, effectively merging their memory into a single twenty-terabyte brain. You just spin up the container, point your computer at another machine, and suddenly you are running massive models that would normally crash a standard laptop. This is genuinely useful because it turns hobbyist hardware into a serious server cluster without needing expensive enterprise gear. Just grab the toolbox, link the nodes, and watch your local AI run circles around the competition.

🆔 @hackernewsgithubprojects

📦 qiaoshouqing/drive-escape

Drive Escape Weekend Map

Drive Escape takes the guesswork out of weekend road trips by showing exactly how far you can travel from any city in five hours. You simply type in a city name, and the tool instantly paints a color-coded map revealing every neighborhood reachable within that time limit. Green areas mean a quick drive, while red zones show how far you can actually get before your time runs out. It uses real driving routes instead of straight-line distance, so you know your actual commute to a getaway spot. The app works globally, supports mobile screens, and is completely free to use without paid services.

🆔 @hackernewsgithubprojects

📦 diffusionstudio/editor

The Open-Source Video Editor Built for Coding Agents

Build videos with code instead of clicking a timeline. Diffusion Studio lets AI agents write TSX scripts to generate motion graphics, captions, and entire scenes that appear instantly in an editable editor. You can ask an agent to recreate an animation style or extract a clip from a link, and the code mounts directly into the interface so a human can tweak the result manually. It combines generative creation with precise editing control, making it a fresh way to produce polished video content through plain language instructions.

🆔 @hackernewsgithubprojects

📦 tsz-org/tsz

tsz: The Rust-Powered TypeScript Speedster

tsz is a blazing-fast TypeScript compiler written in Rust that aims to be twice as quick as the current Go version of the checker. It strips away bloat to deliver a lightning-quick type check, making large codebases feel instantaneous. This is the tool developers have been waiting for to cut through compile lag without sacrificing accuracy. It is the future of fast, reliable TypeScript tooling.

🆔 @hackernewsgithubprojects

📦 dahlia/optique

Optique: Type-Safe CLI Parsing for TypeScript

Build command line interfaces that actually enforce your types. Optique lets you combine small parsing pieces into complex tools while TypeScript tracks exactly what options are available. If you pick a local mode, it knows you need a port. If you choose cloud, it expects a provider key. No more guessing which flag belongs to which command because the code structure forces you to handle every case correctly from the start. It even generates help text and autocomplete scripts automatically. Stop writing brittle command line code and start building tools that catch errors before you even run them. Check out Optique to see how composition makes complex CLIs simple.

🆔 @hackernewsgithubprojects

📦 armada-os/armada

Armada: SteamOS for Handhelds

Armada turns ARM handhelds into a SteamOS experience without needing a PC. It boots from an SD card and lets you install Linux alongside Android, keeping your games and system separate. The interface looks just like Steam, but you can tweak power, fan curves, and compatibility settings right from the controller. It even updates itself over the air, so you never have to redownload anything. The catch is that it is still experimental, so flashing it requires careful steps that could brick your device. If you want a desktop-grade gaming system on a small handheld, this is the open-source path to get there, but proceed with caution.

🆔 @hackernewsgithubprojects

📦 isaqueseneda/shieldfont

ShieldFont: The Typeface That Fools AI Scrapers

ShieldFont is the typeface that finally flips the AI scraping script by making your content useless to mass data harvesters. Here is the clever part: it swaps your words for grammatically correct decoys that look totally normal to a human but confuse automated bots. A custom font then restores your original text on your screen, while the raw HTML source fed to AI models contains only those nonsensical swaps. It raises the cost of stealing your writing without breaking the page for real readers. It is a brilliant, defensive tool for writers who want to keep their words theirs. Give your writing the protection it deserves.

🆔 @hackernewsgithubprojects

📦 virtudude/armada

Armada: SteamOS for Handhelds

Armada is the handheld gaming distro that brings the polished Steam Deck experience to ARM-based handhelds. Instead of a locked-down console, it offers a full Linux desktop with KDE, letting you tweak power, fan, and game settings directly from the Steam UI. It even supports over-the-air updates, so you can patch the system without reflashing your card. The catch is that it is still experimental prototype software, so flashing the bootloader carries a real risk of bricking your device. If you want a powerful, customizable gaming PC in your pocket and are brave enough to tinker, this is your playground.

🆔 @hackernewsgithubprojects

📦 mengto/kage

Kage: A Night Walk Through a Kyoto Temple

Kage lets you scroll through a live 3D night walk inside a Kyoto mountain temple, where the camera glides past lanterns and rain without loading new pages. It builds the entire scene—temple, trees, and moon—right in your browser using JavaScript, so you see wet leaves and fog reacting as you move. The experience feels like flipping through an art book, mixing those moving visuals with still photos for depth. No app download or heavy setup is needed. It shows how much atmosphere you can create with just one file. It is a beautiful, quiet piece of web art worth seeing.

🆔 @hackernewsgithubprojects