This media is not supported in your browser
VIEW IN TELEGRAM
π¦ erensrtky/cogit
cogit: Why your AI code exists
Trace every line of AI-generated code back to the exact decision that created it. cogit is a plugin for Claude Code that records every message you send as a thought commit, linking your code changes to the reasoning behind them. This solves the common problem of forgetting why an AI wrote a specific function, especially when juggling multiple projects. It captures the intent, rejected options, and final outcome in real-time, creating a private, local map of your sessions. You can then open any file to see the history of messages that touched it, including the reasons for each change.
π @hackernewsgithubprojects
cogit: Why your AI code exists
Trace every line of AI-generated code back to the exact decision that created it. cogit is a plugin for Claude Code that records every message you send as a thought commit, linking your code changes to the reasoning behind them. This solves the common problem of forgetting why an AI wrote a specific function, especially when juggling multiple projects. It captures the intent, rejected options, and final outcome in real-time, creating a private, local map of your sessions. You can then open any file to see the history of messages that touched it, including the reasons for each change.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ labscommunity/npu-render
Windows Desktop Runs on the NPU
Your entire Windows desktop can be drawn by the tiny neural processing unit hiding in your laptop, bypassing your graphics card completely. This project, called N P U Render, is a research prototype that swaps the standard drawing engine for one running on the N P U. Instead of letting the main graphics card handle the visuals, it captures those instructions and feeds them into the N P U using a special toolkit. It renders real windows like Notepad and the Start menu with exact pixel accuracy.
π° https://news.ycombinator.com/item?id=50242224
π @hackernewsgithubprojects
Windows Desktop Runs on the NPU
Your entire Windows desktop can be drawn by the tiny neural processing unit hiding in your laptop, bypassing your graphics card completely. This project, called N P U Render, is a research prototype that swaps the standard drawing engine for one running on the N P U. Instead of letting the main graphics card handle the visuals, it captures those instructions and feeds them into the N P U using a special toolkit. It renders real windows like Notepad and the Start menu with exact pixel accuracy.
π° https://news.ycombinator.com/item?id=50242224
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ jaredtrichard/grok-factory
Grok Factory: The AI Crew That Ships Code and Research
Grok Factory runs a full digital crew on a single shared computer, where one bot handles everything from writing software to hunting down profitable business ideas. It works like a ship, where you talk only to the Firstmate, who delegates heavy lifting to specialized cloud agents. The most surprising part is its research pipeline, which automatically scans for money-making ideas, builds financial models, and writes investment theses entirely on its own. It treats every task as a pull request, ensuring you review and merge only what is actually useful. This setup solves the problem of juggling multiple complex tasks by giving you a single, organized point of contact.
π @hackernewsgithubprojects
Grok Factory: The AI Crew That Ships Code and Research
Grok Factory runs a full digital crew on a single shared computer, where one bot handles everything from writing software to hunting down profitable business ideas. It works like a ship, where you talk only to the Firstmate, who delegates heavy lifting to specialized cloud agents. The most surprising part is its research pipeline, which automatically scans for money-making ideas, builds financial models, and writes investment theses entirely on its own. It treats every task as a pull request, ensuring you review and merge only what is actually useful. This setup solves the problem of juggling multiple complex tasks by giving you a single, organized point of contact.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ frank-zy-dou/ndm-gym
NDM Gym: Neural World Models for Humanoid Control
NDM Gym is the neural dynamics modeling toolkit that treats world models as full physics simulators. Instead of relying on slow, traditional physics engines for every training step, it trains a lightweight neural network to predict how a humanoid robot moves, allowing controllers to learn complex skills like motion tracking and object interaction much faster. The real surprise is that these neural models can replace the entire physics engine, running thousands of simulation steps per second on a graphics card while staying accurate. If you want to build humanoids that learn from imagination rather than brute-force simulation, this is the benchmark to know.
π @hackernewsgithubprojects
NDM Gym: Neural World Models for Humanoid Control
NDM Gym is the neural dynamics modeling toolkit that treats world models as full physics simulators. Instead of relying on slow, traditional physics engines for every training step, it trains a lightweight neural network to predict how a humanoid robot moves, allowing controllers to learn complex skills like motion tracking and object interaction much faster. The real surprise is that these neural models can replace the entire physics engine, running thousands of simulation steps per second on a graphics card while staying accurate. If you want to build humanoids that learn from imagination rather than brute-force simulation, this is the benchmark to know.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ infosave2007/cmf
One File, Many Models: CMF
A single file now holds the brain, the mouth, and the memory of an artificial intelligence. That is the core promise of CMF, a new standard for local computing. It stops requiring a tangled mess of folders and complex Python environments. Instead, it packs everything into one self-contained package that your computer can read instantly without copying data. This means you can run powerful language models on your laptop or desktop using a simple command line tool. It is fast because it skips the heavy lifting of traditional software frameworks. For developers who want to run AI locally without the headache, this is a significant shift toward simplicity.
π @hackernewsgithubprojects
One File, Many Models: CMF
A single file now holds the brain, the mouth, and the memory of an artificial intelligence. That is the core promise of CMF, a new standard for local computing. It stops requiring a tangled mess of folders and complex Python environments. Instead, it packs everything into one self-contained package that your computer can read instantly without copying data. This means you can run powerful language models on your laptop or desktop using a simple command line tool. It is fast because it skips the heavy lifting of traditional software frameworks. For developers who want to run AI locally without the headache, this is a significant shift toward simplicity.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ jessepinkman9900/claude-second-brain
Claude Second Brain: Your Notes, Compounded
Stop letting your notes rot in folders. Claude Second Brain builds you a personal knowledge base that actually grows smarter over time. You drop in an article or a book, and Claude reads it, extracts the key ideas, and weaves them into a connected wiki of your own notes. It links new facts to old ones, flags when sources disagree, and keeps everything searchable locally. You get cited answers, not just a list of files. It lives in your GitHub and opens in Obsidian, so you own the data and can read it offline. It is the difference between storing information and actually building understanding.
π @hackernewsgithubprojects
Claude Second Brain: Your Notes, Compounded
Stop letting your notes rot in folders. Claude Second Brain builds you a personal knowledge base that actually grows smarter over time. You drop in an article or a book, and Claude reads it, extracts the key ideas, and weaves them into a connected wiki of your own notes. It links new facts to old ones, flags when sources disagree, and keeps everything searchable locally. You get cited answers, not just a list of files. It lives in your GitHub and opens in Obsidian, so you own the data and can read it offline. It is the difference between storing information and actually building understanding.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ pa-ulander/color-buddy
ColorBuddy: Your VS Code Color Companion
Your editor is quietly highlighting every color in your code with tiny, clickable swatches, but only if you install ColorBuddy. It is a VS Code extension that hunts down color definitions hiding in CSS variables, Tailwind classes, or plain text, and shows you exactly what they look like on hover. The coolest part? It calculates accessibility contrast ratios for you, instantly telling you if your text is readable or fails basic standards. No more squinting at hex codes wondering if they pass. It turns messy styling files into a clean, visual palette that saves hours of debugging.
π @hackernewsgithubprojects
ColorBuddy: Your VS Code Color Companion
Your editor is quietly highlighting every color in your code with tiny, clickable swatches, but only if you install ColorBuddy. It is a VS Code extension that hunts down color definitions hiding in CSS variables, Tailwind classes, or plain text, and shows you exactly what they look like on hover. The coolest part? It calculates accessibility contrast ratios for you, instantly telling you if your text is readable or fails basic standards. No more squinting at hex codes wondering if they pass. It turns messy styling files into a clean, visual palette that saves hours of debugging.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ medovina/natural_lean
Natural Lean: Write Math in Plain English
You can write formal mathematical proofs in plain English that a computer actually understands. The library called Natural Lean lets you type theorems and definitions using ordinary sentences instead of rigid programming syntax. When you write a claim like the cancellation law for addition, the tool instantly converts that text into native code. It then checks your logic to confirm the math is correct, right inside your editor. This removes the steep learning curve usually associated with formal verification. You stop fighting with special symbols and start thinking like a mathematician. It is still early in development, but the vision is clear.
π @hackernewsgithubprojects
Natural Lean: Write Math in Plain English
You can write formal mathematical proofs in plain English that a computer actually understands. The library called Natural Lean lets you type theorems and definitions using ordinary sentences instead of rigid programming syntax. When you write a claim like the cancellation law for addition, the tool instantly converts that text into native code. It then checks your logic to confirm the math is correct, right inside your editor. This removes the steep learning curve usually associated with formal verification. You stop fighting with special symbols and start thinking like a mathematician. It is still early in development, but the vision is clear.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ hanxiao/deepseek-v4.1-flash-1017-fp8-abliterated-4x-rtx-pro-6000
DeepSeek V4.1 Flash: The Fastest Local LLM Setup
DeepSeek V4.1 Flash abliterated is the high-performance inference setup that finally makes massive language models feel instant on consumer-grade hardware. It takes a huge, complex AI model and squeezes it onto four mid-tier graphics cards, using clever compression and a specialized engine to stream text at nearly five hundred words per second. The real magic is the speculative decoding trick, which lets the machine guess multiple words at once and verify them in parallel, turning a slow, clunky chatbot into a lightning-fast coding agent that handles enormous documents without breaking a sweat.
π @hackernewsgithubprojects
DeepSeek V4.1 Flash: The Fastest Local LLM Setup
DeepSeek V4.1 Flash abliterated is the high-performance inference setup that finally makes massive language models feel instant on consumer-grade hardware. It takes a huge, complex AI model and squeezes it onto four mid-tier graphics cards, using clever compression and a specialized engine to stream text at nearly five hundred words per second. The real magic is the speculative decoding trick, which lets the machine guess multiple words at once and verify them in parallel, turning a slow, clunky chatbot into a lightning-fast coding agent that handles enormous documents without breaking a sweat.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ workingprototype/fuobe
Fuobe: The Open Creative Studio
Fuobe turns your own graphics card into a full-scale production studio. Forget cloud subscriptions or renting expensive hardware. You run open-weight models locally on your graphics card through ComfyUI. The real surprise is the scale of what it ships out of the box. You get over a hundred camera motion presets, hundreds of visual effects, and more than a hundred one-click apps for portraits and product shots. It is not just an image generator. It is a complete editing suite for video, three-dimensional assets, and audio. If you want total creative control without the recurring fees, this is your studio.
π @hackernewsgithubprojects
Fuobe: The Open Creative Studio
Fuobe turns your own graphics card into a full-scale production studio. Forget cloud subscriptions or renting expensive hardware. You run open-weight models locally on your graphics card through ComfyUI. The real surprise is the scale of what it ships out of the box. You get over a hundred camera motion presets, hundreds of visual effects, and more than a hundred one-click apps for portraits and product shots. It is not just an image generator. It is a complete editing suite for video, three-dimensional assets, and audio. If you want total creative control without the recurring fees, this is your studio.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ google-deepmind/code-verification
Code Verification: Proving Signal is Secure
Code Verification is the security proof that verifies Signal's machine code with mathematical certainty. Google Deep Mind used artificial intelligence to generate formal proofs that the compiled code actually works exactly as intended, catching any hidden bugs or backdoors. It checks the actual low level instructions your phone runs, not just the high level design documents. This gives you independent, ironclad assurance that your private messages stay secure. No more guessing if the app is safe. You can now trust the math. That is the power of verified software.
π @hackernewsgithubprojects
Code Verification: Proving Signal is Secure
Code Verification is the security proof that verifies Signal's machine code with mathematical certainty. Google Deep Mind used artificial intelligence to generate formal proofs that the compiled code actually works exactly as intended, catching any hidden bugs or backdoors. It checks the actual low level instructions your phone runs, not just the high level design documents. This gives you independent, ironclad assurance that your private messages stay secure. No more guessing if the app is safe. You can now trust the math. That is the power of verified software.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ torsher-app/torsher
Torsher: The All-In-One Media Server
Torsher is a self-hosted media center that lives entirely inside a single Docker container, designed specifically for watching on a television rather than a computer screen. The most surprising capability is that you can start playing a movie or show immediately while the file is still downloading in the background, which means you never wait for a full download to finish before pressing play. This one tool replaces a complicated stack of separate applications for managing torrents, storing files, and organizing your library, keeping everything in one place with a simple remote-control interface.
π° https://news.ycombinator.com/item?id=50253432
π @hackernewsgithubprojects
Torsher: The All-In-One Media Server
Torsher is a self-hosted media center that lives entirely inside a single Docker container, designed specifically for watching on a television rather than a computer screen. The most surprising capability is that you can start playing a movie or show immediately while the file is still downloading in the background, which means you never wait for a full download to finish before pressing play. This one tool replaces a complicated stack of separate applications for managing torrents, storing files, and organizing your library, keeping everything in one place with a simple remote-control interface.
π° https://news.ycombinator.com/item?id=50253432
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ everruns/everruns
Everruns: The Engine for AI Agents That Never Forget
Run complex AI agents that actually finish their jobs, even when they take hours. Everruns solves the annoying problem where an agent crashes or loses its memory mid-task, starting everything over from scratch. It works like a durable checkpoint system, saving the agent's progress so it can pick up right where it left off without losing any context. This is built in Rust for serious reliability, letting you scale agents across different AI providers without the usual headaches. If you want your AI to handle long, multi-step tasks that actually stick around, this is the engine you need.
π @hackernewsgithubprojects
Everruns: The Engine for AI Agents That Never Forget
Run complex AI agents that actually finish their jobs, even when they take hours. Everruns solves the annoying problem where an agent crashes or loses its memory mid-task, starting everything over from scratch. It works like a durable checkpoint system, saving the agent's progress so it can pick up right where it left off without losing any context. This is built in Rust for serious reliability, letting you scale agents across different AI providers without the usual headaches. If you want your AI to handle long, multi-step tasks that actually stick around, this is the engine you need.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ luwudynamics/xgoduck_runtime_arduino
XGO Duck: A Tiny Robot That Walks
This tiny duck robot actually picks things up using real-time neural networks. It runs on an Arduino Uno Q, where a tiny chip handles the legs while a bigger chip runs the brain. The cool part? If the duck falls over, the software detects it and automatically triggers a recovery model to get it back on its feet. This project bridges the gap between hobbyist hardware and serious robotics. You get a web interface to control its walk, head, and even its mouth. It is a fantastic example of how efficient hardware can handle complex behavior without a supercomputer.
π @hackernewsgithubprojects
XGO Duck: A Tiny Robot That Walks
This tiny duck robot actually picks things up using real-time neural networks. It runs on an Arduino Uno Q, where a tiny chip handles the legs while a bigger chip runs the brain. The cool part? If the duck falls over, the software detects it and automatically triggers a recovery model to get it back on its feet. This project bridges the gap between hobbyist hardware and serious robotics. You get a web interface to control its walk, head, and even its mouth. It is a fantastic example of how efficient hardware can handle complex behavior without a supercomputer.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ velo4705/awesome-free-byok-models
Awesome Free BYOK Models: No Credit Card Needed
This list only includes AI model providers where you can grab an API key without ever touching a credit card. It solves the annoying problem of fake free trials that quietly expire or demand payment details upfront. Every entry is tested by real people to confirm it stays free and works instantly. You get a simple table showing exactly which models run, how much data they handle, and how fast they respond. No hidden costs, no surprise bills, just verified access. It cuts through the noise so you can find a reliable, zero-cost API that actually works for your projects.
π @hackernewsgithubprojects
Awesome Free BYOK Models: No Credit Card Needed
This list only includes AI model providers where you can grab an API key without ever touching a credit card. It solves the annoying problem of fake free trials that quietly expire or demand payment details upfront. Every entry is tested by real people to confirm it stays free and works instantly. You get a simple table showing exactly which models run, how much data they handle, and how fast they respond. No hidden costs, no surprise bills, just verified access. It cuts through the noise so you can find a reliable, zero-cost API that actually works for your projects.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ ichris4/wiicompiled_vr
Mario Kart Wii Runs Natively on PC
WiiCompiled VR is the native port that finally lets you race Mario Kart Wii on your PC without a single line of emulated code. Instead of faking a Wii console with an interpreter, the project statically recompiles the gameβs original instructions into native C++ that your computer actually executes. The standout feature is experimental virtual reality support, where you sit directly in the cockpit. Your steering wheel turns in sync with your tracked controllers, creating a genuine sense of driving rather than just looking at a screen.
π @hackernewsgithubprojects
Mario Kart Wii Runs Natively on PC
WiiCompiled VR is the native port that finally lets you race Mario Kart Wii on your PC without a single line of emulated code. Instead of faking a Wii console with an interpreter, the project statically recompiles the gameβs original instructions into native C++ that your computer actually executes. The standout feature is experimental virtual reality support, where you sit directly in the cockpit. Your steering wheel turns in sync with your tracked controllers, creating a genuine sense of driving rather than just looking at a screen.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ bornholm/indecis
indecis
indecis lets you build tiny decision models that answer specific questions about text with calibrated probabilities instead of generating free-flowing replies. It takes a small pretrained encoder and fully fine-tunes it on your own labeled examples so it can classify topics, detect intent, or score urgency in milliseconds. The whole thing runs on a standard laptop without a GPU, using less memory than most browser tabs and training in just thirty minutes on twenty thousand examples. If you need fast, reliable, and private text classification, this is a surprisingly capable tool. It proves you don't need a massive AI infrastructure to get smart, structured answers from your data.
π @hackernewsgithubprojects
indecis
indecis lets you build tiny decision models that answer specific questions about text with calibrated probabilities instead of generating free-flowing replies. It takes a small pretrained encoder and fully fine-tunes it on your own labeled examples so it can classify topics, detect intent, or score urgency in milliseconds. The whole thing runs on a standard laptop without a GPU, using less memory than most browser tabs and training in just thirty minutes on twenty thousand examples. If you need fast, reliable, and private text classification, this is a surprisingly capable tool. It proves you don't need a massive AI infrastructure to get smart, structured answers from your data.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ bladedevoff/stuntd
stuntd: The AI Proxy That Learns Your App
stuntd acts as a smart middleman that watches how your app uses large language models to make specific decisions, like sorting tickets or checking if a command is safe. Instead of paying for every single question, it quietly records those answers and trains a tiny, lightweight model to replicate them locally on your own hardware. This means it can answer the same questions in about twenty milliseconds without ever sending your data to an outside server. If it is unsure, it safely hands the job back to the main provider, ensuring accuracy while cutting down on costs and latency.
π @hackernewsgithubprojects
stuntd: The AI Proxy That Learns Your App
stuntd acts as a smart middleman that watches how your app uses large language models to make specific decisions, like sorting tickets or checking if a command is safe. Instead of paying for every single question, it quietly records those answers and trains a tiny, lightweight model to replicate them locally on your own hardware. This means it can answer the same questions in about twenty milliseconds without ever sending your data to an outside server. If it is unsure, it safely hands the job back to the main provider, ensuring accuracy while cutting down on costs and latency.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ leonxlnx/verdant-forest
Verdant Forest
Verdant Forest is the browser-based woodland that renders thousands of unique trees in real time. This project builds a dense, temperate forest directly in your web browser using actual curved geometry rather than simple flat images. You can fly through the scene with your keyboard to explore thousands of individual grass blades, ferns, and wildflowers. The real surprise here is the scale. The engine places over seven thousand distinct trees and millions of small ground details, all moving and reacting to wind simultaneously. It solves the problem of digital nature feeling flat by creating a living, breathing environment that you can actually walk through.
π @hackernewsgithubprojects
Verdant Forest
Verdant Forest is the browser-based woodland that renders thousands of unique trees in real time. This project builds a dense, temperate forest directly in your web browser using actual curved geometry rather than simple flat images. You can fly through the scene with your keyboard to explore thousands of individual grass blades, ferns, and wildflowers. The real surprise here is the scale. The engine places over seven thousand distinct trees and millions of small ground details, all moving and reacting to wind simultaneously. It solves the problem of digital nature feeling flat by creating a living, breathing environment that you can actually walk through.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ rhfeiyang/geb
GEB: The AI That Remembers Objects, Not Just Time
GEB is the long-video memory framework that finally stops confusing two identical mugs. Instead of just listing events in order, it builds a distinct biography for every single physical object it sees. This means the system knows exactly which red cup went into the dishwasher, even if another identical one stayed on the counter. By linking visual sightings across hours, it tracks the actual history of specific items rather than just generic descriptions. Itβs a clever way to give machines real-world common sense about identity. If you work with video data, this is the kind of detail that actually matters.
π @hackernewsgithubprojects
GEB: The AI That Remembers Objects, Not Just Time
GEB is the long-video memory framework that finally stops confusing two identical mugs. Instead of just listing events in order, it builds a distinct biography for every single physical object it sees. This means the system knows exactly which red cup went into the dishwasher, even if another identical one stayed on the counter. By linking visual sightings across hours, it tracks the actual history of specific items rather than just generic descriptions. Itβs a clever way to give machines real-world common sense about identity. If you work with video data, this is the kind of detail that actually matters.
π @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
π¦ leonvanzyl/cubefarm
CubeFarm: The 3D Office for AI Developers
CubeFarm is the cartoon three-dimensional office that finally lets you watch your code get written. It turns a list of tasks into a living building where a team of artificial intelligence workers sits at desks, coding through your project issues. The most surprising part is that you can literally walk up to their monitors to see their real terminals running, or watch the quality assurance testers open your actual web app in a browser to verify the fixes. It is a fun, transparent way to manage automated software development without staring at a wall of text.
π @hackernewsgithubprojects
CubeFarm: The 3D Office for AI Developers
CubeFarm is the cartoon three-dimensional office that finally lets you watch your code get written. It turns a list of tasks into a living building where a team of artificial intelligence workers sits at desks, coding through your project issues. The most surprising part is that you can literally walk up to their monitors to see their real terminals running, or watch the quality assurance testers open your actual web app in a browser to verify the fixes. It is a fun, transparent way to manage automated software development without staring at a wall of text.
π @hackernewsgithubprojects