298 subscribers
14K videos
14.6K links
Download Telegram
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ dev-armaan/satellites

Watched: See every public satellite image of Earth

The thing that makes Watched feel like magic is that the entire orbit calculation happens inside your web browser. This free tool takes any place on Earth and replays the last seven days of satellite motion to show you exactly which spacecraft flew over and which real photos they captured. It even animates the camera sweep, revealing the image line by line just like the satellite recorded it. No sign-up required, and the math is handled locally by your device. If you have ever wondered who was looking at your city from space this week, this is the only place you need to go.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ adventists-ai/sophon-realtime

Sophon-Realtime: The voice model that hears your tone

Sophon-Realtime is the open voice model that actually hears your tone while speaking with its own. Most AI assistants just wait for you to stop talking, but this one listens continuously, catching your sarcasm, sighs, and even your breaths in real time. It doesn't just read text back to you; it decides on its own when to laugh, clear its throat, or stress a word to match the mood of the conversation. Because it runs entirely on a single graphics card, you can host this expressive, interrupt-capable assistant for free without relying on closed cloud services.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ roboflow/re-id

Re-ID: The Python Tool for Person Re-Identification

Roboflow’s Re-ID library turns any snapshot of a person into a simple numeric fingerprint that lets computers recognize them later, even if they are wearing different clothes or standing under a different camera. It solves the difficult problem of linking separate video clips or camera feeds to follow the same individual without needing to track them in real time. You feed it a photo or a video frame, and it compares that image against others using a mathematical distance check. This is useful for security systems, sports analytics, and crowd management where you need to know who is who across different locations.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ skild-ai/steel

Steel: The Physics Engine Built for Robots

Steel calculates how robots move and collide, but it does so at speeds that would feel like cheating. Built by Skild and NVIDIA, this solver runs three to thirteen times faster than the industry standard, PhysX, while actually using less memory on your graphics card. It was designed specifically for the messy reality of robot learning, where you often have to simulate many different robots performing different tasks at the same time. Instead of slowing down when the mix gets complicated, it keeps running at full speed. This matters because training smart robots requires billions of simulated hours, and being faster means you get results in days instead of months.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ shenh10/huggingarch

HuggingArch: The Model Architect

HuggingArch is the planning tool that tells you exactly how much your model costs before you spend a dime on hardware. It reads a model's configuration file and calculates its memory footprint, speed limits, and data movement without ever downloading the heavy weights. This is brilliant because you can see if a giant AI will actually fit on your graphics card or run at a snappy speed, purely by looking at the math. It turns a blind guess into a solid number. Now you know the hard limits before you start your deployment. Stop guessing and start calculating.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ cywang735/video-coder

How AI Codes Your Videos

Video-Coder is the coding agent that builds video scenes by writing Blender code instead of guessing pixels. Most video tools struggle to keep objects moving correctly because they lack a deep understanding of physics. This project solves that problem by having an AI model write Python scripts that explicitly place objects, set camera angles, and animate movements in a 3D environment. The system runs this code to create a draft video, which then guides a separate generator to produce the final footage. By using this code-driven approach, the result has accurate structure and motion.

πŸ†” @hackernewsgithubprojects
Media is too big
VIEW IN TELEGRAM
πŸ“¦ neuronto/agentic-resource-discovery

Agentic Resource Discovery: The Search Engine for AI Tools

Agentic Resource Discovery acts as a live search engine for AI capabilities, but the surprising part is that it verifies every single tool by actually connecting to the server and reading its code. Most directories just trust the text publishers write, which means dead links and broken features clutter your results. This project fixes that by only listing tools that are currently online and working, giving you thirty-two thousand verified options instead of guesses. It essentially turns the chaotic world of scattered AI plugins into one clean, searchable list where you can find exactly what you need without manually wiring up every connection.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ ohmzi/sure

Sure: The Finance App Built With AI

Sure is the personal finance app that treats every user as a co-developer. Built on a Ruby and JavaScript foundation, it connects to dozens of banks, investment platforms, and crypto exchanges to give you a clear view of your money. The truly surprising part is how it handles the messy work: instead of just listing transactions, the app uses AI to automatically review bills, categorize spending, and even suggest smart configurations for recurring payments. It’s less like a static ledger and more like a financial assistant that actually organizes your life.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ onlyreadai/onlyread

Onlyread: Stop AI Agents From Deleting Your Data

Onlyread stops AI browser agents from accidentally deleting or posting on your behalf. When you let an AI tool handle your email or social media, it runs in your actual logged-in account, creating a huge risk if it misunderstands an instruction or gets tricked by a malicious page. This Chrome extension enforces a strict read-only mode at the network level, blocking any request to send, edit, or delete before it leaves your browser. It works with any AI agent in Chrome, meaning you can trust the software to read your inbox or research people on LinkedIn without ever worrying about accidental damage.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ cd1s/enginex-os

EngineX OS: The DJ Hack You Didn't Know You Needed

EngineX OS turns a standard Denon DJ controller into a four-deck powerhouse with a simple USB flash. The most surprising part is how it unlocks professional-grade track splitting directly on the hardware. You can now isolate vocals, melody, bass, and drums in real time without needing a computer. This solves the problem of bulky gear by letting you create complex transitions and mashups right on your deck. It also connects to streaming services like Spotify and Apple Music, so you can play your whole library instantly. This is a powerful tool for any DJ who wants more flexibility without buying expensive new equipment.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ twili8/homer

Homer: The Self-Defending Windows Loader

Homer rewrites the rules for how Windows code loads into memory by destroying the fixed patterns that scanners look for. It takes a standard Windows program and wraps it in a small, randomized shell that hides its identity every single time you build it. Instead of having a static signature that analysts can easily spot, the loader uses a fresh secret and shuffled layout for each run, forcing defenders to understand the code rather than just matching bytes. This makes static analysis incredibly difficult without needing to execute the file.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ davide-troiani/moeshift

Run 125B AI Locally on Your Mac

MoEShift is the Apple Silicon tool that finally lets you run massive AI models without crashing your laptop. While other tools choke on memory, this project uses a clever trick to keep your Mac responsive while it chugs through twelve and a half billion parameters. It runs the Qwen 3.8 model in under ten gigabytes of RAM, which means you can actually browse the web or write code while it thinks. The magic happens on your fast SSD, not just your memory, so it stays smooth even on a modest machine. If you want serious local AI power without buying a server, this is the one to watch.

πŸ“° https://news.ycombinator.com/item?id=50358319

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ sshah03/recalc

recalc turns a spreadsheet into a live dashboard for your AI agents

recalc turns a standard spreadsheet into a live dashboard where every single cell can be a terminal window running an artificial intelligence agent. You can set up a grid where an agent writes its code in one cell and a test script in the cell right next to it. If the agent claims the code works but the test fails, the cell turns red instantly so you can spot the lie before it ships. It is built for people who use coding agents but hate losing track of what they are doing. Instead of juggling five different chat windows, you watch them all work on one shared screen.

πŸ“° https://news.ycombinator.com/item?id=50413970

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ shuttle-browser/shuttle

Shuttle: The Browser Built for AI Agents

Shuttle is a stripped-down Chromium fork that gives every artificial intelligence agent its own logged-in identity and its own set of tabs right inside your regular browser window. It solves a massive headache where agents used to steal your cookies or get locked out of your accounts entirely. Now, each AI gets its own persistent storage and a built-in connection layer, so it stays logged in and works in total isolation. You can watch the agent browse, grab a tab, or kick it out of your session without any extra software. It is the first browser that actually treats AI agents as first-class citizens.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ yihaohu0118/viss

ViSS: The Video Model That Decides What To Watch

A vision language model can now decide exactly which video frames to inspect before answering a question. This project, called Viss, separates what a model knows from how it looks at the evidence. Instead of forcing the system to process every single frame of a long video, it starts with the recent moments and only fetches specific past clips if it is unsure. This saves massive computing power while keeping the model accurate. It is a fresh approach to making streaming video understanding efficient and reliable. If you are building systems that watch live feeds, this is the one to study.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ pmndrs/klipp

Klipp: The Cinemachine for the Web

Direct your own cinematic shots in any browser-based three-dimensional scene. Klipp brings the power of game engine virtual cameras to the web, letting you define multiple camera angles with specific priorities. When you switch between them, it doesn't just jump; it blends the position and rotation smoothly, creating those high-end transitions seen in AAA titles. Think of it as a director for your code, handling the tricky math of following targets and framing groups so your camera feels alive instead of static. It runs on pure math, so it works with any renderer you choose.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ cx-artlab/agent-robot-avatar

Give Your AI a Face With Agent Robot Avatar

Drop a living robot face into your web app that actually feels alive. Agent Robot Avatar is a tiny piece of code that turns a blank space into a character with eyes that follow your mouse and a body that squishes when you poke it. The coolest part is how it reacts to your AI’s work. When your chatbot is thinking, the robot looks away and waits. When it finishes, it celebrates or looks sad if something went wrong. It is built with simple web tools and no heavy software, so it loads instantly.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ shawnthebeachy/return-affinity-colors

Get Your Affinity Colors Back

Affinity Designer for Version Three launched with every tool icon stripped of its color, leaving your interface feeling stark and empty. A developer named Shawn The Beachy built a tool called Return Affinity Colors to fix this. It works by digging into the app's core files and swapping those blank gray symbols for the vibrant icons from the previous version. You can even dump the original graphics, tweak them, and import your own custom set. It is a simple command line utility that restores the visual charm many users missed.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ qinggo/engram-peft

Engram-PEFT: The 240x Memory Trick for AI

Engram-PEFT lets you inject massive amounts of knowledge into large language models without slowing them down at all. This unofficial implementation of the DeepSeek Engram paper uses a clever trick called sparse retrieval to add a high-capacity memory layer that sits alongside the model. The magic is that it only activates a tiny fraction of that memory during each step of inference, meaning you get huge storage for facts without the usual computational overhead. It is essentially a way to give an AI a giant library of specific data without making it think slower.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ uclanlp/openvlharness

OpenVLHarness: The Visual Reasoning Upgrade

OpenVLHarness is the visual reasoning harness that turns any vision model into a detective. It works by giving the model access to a toolkit of specialized services, like a zoom lens, a text reader, and even a web browser. This allows the AI to look closer at details, read signs, and search for context online before answering. The real magic happens in its memory system. Unlike standard models that forget previous steps, this system keeps track of everything it sees, like boxes and images, so it can build on its own discoveries. The result is a massive jump in accuracy without retraining the model.

πŸ†” @hackernewsgithubprojects