📦 pinto0309/ocec

OCEC: The 112 KB Eye Detector

The tiniest eye detector in the wild is a mere one hundred and twelve kilobytes, yet it judges if eyes are open or closed in less than a millisecond. Named OCEC, this ultra-lightweight model solves a specific problem: most computer vision tools are too heavy for real-time tracking on low-power devices. By analyzing tiny twenty-four by forty pixel crops, the project provides a training pipeline and several model sizes that balance speed and accuracy. This is useful for anyone building responsive interfaces or accessibility tools where processing power is limited.

🆔 @hackernewsgithubprojects

📦 pulse-energy-eu/slopshape

SlopShape: How to Spot AI Writing

SlopShape lets you spot AI-written business articles without reading a single word. Instead of looking for specific keywords or grammar errors, this research project measures the hidden structural habits of text. The team compared thousands of real human blogs against AI versions written by five different models. They found that AI posts follow predictable patterns in how they organize ideas, which remain visible even if you rewrite the sentences. This tool creates a reliable way to verify content authenticity on the web. If you rely on online articles for business decisions, knowing how to filter out machine-generated filler is becoming an essential skill for staying informed.

🆔 @hackernewsgithubprojects

📦 angerybob/s2-moe

How S2 MoE Makes AI Faster on Small Chips

S2 MoE is the inference engine that finally makes large AI models run fast on small, power-hungry chips. It solves a major bottleneck where running these complex models on devices like the Jetson Orin is usually too slow because of limited memory. The key trick is a smart guessing game where the model predicts its own next words, but only keeps the guesses that actually save time. By carefully choosing which parts of the model to use, it avoids wasting energy on useless calculations. This means you get speeds up to five times faster than the standard approach.

🆔 @hackernewsgithubprojects

📦 androoagi/starnet

Starnet: A Pixel Art Space Station for Real AI Agents

Your AI agents are no longer confined to a chat window. In StarNet, they live in a pixel art space station that mirrors their actual activity in real time. Every room represents a specific task, and every object you place grants a concrete permission. This visual layout is not decoration; it is the exact workflow your agents execute. You bring your own model keys, so your data stays local, while the agents perform genuine work with real tools and costs. You can run multiple specialized agents concurrently, each with its own workspace. The station acts as a live dashboard for your digital crew.

🆔 @hackernewsgithubprojects

📦 alban1911/rose

Rose: How to Get Every Skin in League of Legends for Free

You can play League of Legends with every single skin in the game without spending a cent, and it is all thanks to an open-source tool called Rose. It works by quietly running in the background, watching your champion select screen, and then swapping in the skin you picked the exact moment the game loads. The best part is not just seeing the skins, but the fact that it lets you share them with your friends in real-time through a cloud relay. Your entire lobby can see the same custom models simultaneously, turning your match into a coordinated fashion show.

🆔 @hackernewsgithubprojects

📦 codezeno/claude-code-usage-monitor

Claude Code Usage Monitor: Never Hit Your AI Limit Again

Claude Code Usage Monitor is the Windows taskbar widget that finally tells you exactly when your AI limits reset. If you live in coding agents, you know the panic of hitting a wall mid-task, but this little tool sits quietly in your system tray, reading your local logins to track usage for Claude, Codex, Cursor, and Grok without sending your data to a single third-party server. The coolest part is that it counts down your remaining allowance in real time, so you can pace your prompts like a pro gamer managing stamina. It is completely free, open source, and works on Windows ten or eleven.

🆔 @hackernewsgithubprojects

📦 mvschwarz/openrig

Openrig: Two AIs, One Brain

Openrig is the developer tool that makes two different AIs work as one team. It runs Claude Code and Codex side-by-side, coordinating them to handle complex coding tasks together. This solves the problem of AI agents working in silos, creating a unified system that can plan, build, and review code in a single workflow. If you want to see how multiple AIs can actually collaborate instead of just chatting, this is the project to watch.

🆔 @hackernewsgithubprojects

📦 danielravina/stemkit

StemKit Splits Songs Into Stems Locally

StemKit lets you pull every instrument out of a YouTube song and play them back individually, all without sending your files to the cloud. This desktop app runs entirely on your computer, meaning your private music library never leaves your machine, and you do not need any special accounts or internet keys to start using it. You simply search for a track or paste a link, and the software uses local processing to separate the vocals, drums, and bass into distinct tracks.

🆔 @hackernewsgithubprojects

📦 derv82/wifit3

WiFi T3: The Cross-Platform USB WiFi Tool

Wifit3 is the cross-platform tool that runs entirely from a USB dongle. It turns a small plug-in radio into a powerful scanner and cracker for local networks. The system talks directly to the hardware, so it works on Windows, macOS, and Linux without needing special drivers. This design removes the usual friction of setting up complex capture tools. You simply plug in the device, and the software handles the rest. It supports a wide range of common chips found in cheap USB adapters. This makes it a practical choice for anyone who wants to audit their own wireless setup.

🆔 @hackernewsgithubprojects

📦 ilkerzg/storycast

Storycast turns any topic into an animated movie

Turn any topic into a fully animated short film with Storycast. Just type in a subject, pick a character, and the platform handles everything from writing the script to animating the visuals and adding the voiceover. It creates a complete, narrated movie in various artistic styles and languages right in your browser. This is the perfect tool for anyone who wants to explain complex ideas simply or just make cool videos without hiring a team. It is a surprisingly simple way to turn plain text into engaging visual stories.

🆔 @hackernewsgithubprojects

📦 liuziyu77/valen

Valen: The AI That Decides Faster Than It Thinks

Valen lets you train a multimodal model that makes split-second decisions by looking at text, images, and video. It solves the problem of slow, overthinking AIs that take too long to answer simple questions. This is a System One model, meaning it reacts instantly rather than reasoning through every step. It’s built on a Qwen backbone but uses a shared decision head to score options directly without generating long answers. The coolest part? It solved a Sokoban puzzle in just over one second, while a much larger model took nearly three minutes to fail. It’s a tiny, fast brain for real-time software.

📰 https://news.ycombinator.com/item?id=49838068

🆔 @hackernewsgithubprojects

📦 tinyhumansai/tinyagents

TinyAgents: The Rust Agent Harness

TinyAgents is the Rust library that finally lets you build durable, resumable AI agents without dragging a Python runtime into your stack. It provides a typed state graph engine where every step of your AI’s logic is checked, saved, and replayable, meaning if a long task crashes or gets interrupted, the system picks up exactly where it left off instead of starting over. You define nodes and edges in standard Rust, and the library handles the heavy lifting of tracking progress, managing checkpoints, and coordinating multiple agents working in parallel.

📰 https://news.ycombinator.com/item?id=48727680

🆔 @hackernewsgithubprojects

📦 ebibibi/ebi-agent-chat-relay

Ebi Agent Chat Relay: Chat with AI Agents in Discord

Ebi Agent Chat Relay is the open-source framework that lets you command powerful AI coding agents directly from your favorite chat apps like Discord or Microsoft Teams. Instead of juggling different interfaces, this tool acts as a universal bridge, turning your casual chat threads into fully functional workspaces where you can run Claude Code, Codex, or even local models. The most impressive part is how it treats every conversation thread as a persistent session, meaning you can start a complex task, step away, and pick up exactly where you left off without losing context.

🆔 @hackernewsgithubprojects

📦 milanofthe/sanity

Sanity: See Your Codebase Live

Sanity turns your entire codebase into a single, zoomable map where you can watch files change in real time. As an AI agent or human edits your project, every alteration flashes on screen, letting you instantly see which parts of the system are moving and why. It also plays back your version history like a movie, so you can step through past changes without opening a terminal. This visual monitor keeps you in the loop while automated tools work, ensuring you never lose track of what is actually happening in your repository. It is the perfect way to stay in control of complex coding tasks.

🆔 @hackernewsgithubprojects

📦 burnigtm/jev-mcp

Jev MCP: The Fast Judge for AI Coding Loops

Jev MCP lets AI coding agents make fast, typed decisions without spinning up expensive generative models for every single step. Instead of asking a large language model to write code for simple checks, this local server connects to TypeSafe Jev to answer specific questions like whether a patch is safe or which tool to use next. It fits right into your workflow, saving you from unnecessary model calls while keeping the decision logic transparent and reviewable. The real takeaway is that you can automate the boring, repetitive parts of your coding loop with confidence, letting the expensive AI do only the creative heavy lifting where it actually matters.

🆔 @hackernewsgithubprojects

📦 tenurehq/tenure

Tenure: The Memory Layer That Keeps Your AI Honest

Tenure is the local memory layer that finally stops your AI tools from forgetting what you already decided. It sits quietly between your chat apps and the model, remembering things like your favorite code style or a database choice you made weeks ago. The coolest part is that it keeps these memories strictly separated by project, so your work code never bleeds into your personal writing. It also keeps a full audit trail, so if the AI acts weird, you can look back and see exactly which memory caused it. You keep total control over what it learns, and everything stays on your machine.

🆔 @hackernewsgithubprojects

📦 deepopen-com/deepopen

DeepOpen: The Fastest Way to Classify Text

DeepOpen is the decision engine that finally kills the slow, expensive guesswork of traditional language models. Instead of writing out an answer word by word, it looks at your text once and instantly spits out the correct category. It handles over a hundred languages in a single pass, taking only thirty-three milliseconds to figure out what a customer actually wants. Because it never generates free text, it cannot hallucinate nonsense, which means you get reliable answers without paying per token. If you are building systems that need to sort tickets or route requests fast, this open-source tool is the smart shortcut you have been waiting for.

🆔 @hackernewsgithubprojects

📦 yetone/magpie

Magpie: One Place to Pick Every Agent's Model

Point any coding agent at any model from a single menu bar app. Magpie solves the headache of juggling different configuration files for tools like Codex, Claude Code, or Gemini. It acts as a local bridge that translates requests between different AI providers, letting you swap models with one click. The coolest part is sharing your existing subscriptions. Sign in to one tool, and Magpie exposes those models to every other agent on your machine. No copying keys, no extra setup. Just pick a model from one list and use it everywhere. It keeps your setup clean and your options open. Stop managing ten different config files.

🆔 @hackernewsgithubprojects

📦 techwolf-ai/ai-first-toolkit

AI-First Toolkit: The Code Auditor

A single install command can score your entire codebase against nine specific design principles for AI compatibility. The AI-First Toolkit is an open-source collection of agent skills that audits, re-engineers, or bootstraps your projects to work effectively with modern artificial intelligence tools. It solves the common problem of legacy code structures that confuse or limit AI coding assistants. By restructuring your project layout and documentation, this toolkit ensures your AI tools have the clear context they need to be truly helpful. It turns a chaotic codebase into a structured environment that agents can understand and navigate with precision.

🆔 @hackernewsgithubprojects

📦 receptron/laya

Laya: AI Decisions in Node

Laya brings instant, probabilistic decision-making directly into your Node.js and TypeScript applications without needing Python or heavy AI libraries. It acts as a fast, brain-like processor that reads a piece of data, like a customer ticket, and answers multiple questions at once with confidence scores. You can ask it to pick a department, rate urgency, or predict risk, and it returns precise numbers in a single pass. This tool runs on standard hardware via ONNX, making advanced AI logic accessible and portable for any developer. If you want to add smart, explainable decisions to your backend without the setup headache, Laya is the way to go.

🆔 @hackernewsgithubprojects

📦 langgenius/mosoo

What Mosoo Actually Is

A developer can publish one coding agent once, and anyone can run it in a secure cloud sandbox without setting up their own infrastructure. This is the core function of Mosoo, an open source gateway that lets you run powerful coding tools like Codex and Claude through a simple web address. It solves the messy problem of managing complex agent setups by handling the isolated execution environment for you, so you do not need to build that layer from scratch. You get a clean, shared place to start, track, and stop agent work, while keeping your project files and conversation history safe across different sessions.

🆔 @hackernewsgithubprojects