📦 jd-opensource/joyai-video-edit

Edit Live Video with JoyAI Video Edit

Watch JoyAI Video Edit transform a live video stream in real time using just a text instruction. Instead of waiting for an entire clip to finish or processing it in slow batches, this system edits frames the moment they arrive. It combines a large language model to understand your request with a diffusion engine that paints new pixels on the fly. The result feels like magic, letting you swap backgrounds, change clothing styles, or remove objects as you watch the screen. It achieves thirty frames per second, turning video editing from a tedious, offline chore into an instant, interactive experience that feels surprisingly natural.

🆔 @hackernewsgithubprojects

📦 xiaobin-rong/unipase

UniPASE Speech Enhancement

UniPASE transforms noisy audio into crystal-clear speech without adding weird robotic artifacts, a rare feat in generative models. This repository offers the official implementation of a state-of-the-art system designed to clean up voice recordings while keeping them sounding natural and authentic. It works by stripping away background noise and intelligently reconstructing missing sound details rather than just deleting bad parts. You get ready-to-use checkpoints that handle everything from basic clarity to fixing dropped audio packets. The setup is straightforward, requiring only standard Python libraries like PyTorch and SciPy to run inference on your own files.

🆔 @hackernewsgithubprojects

📦 fzkuji/gui-agent-harness

GUI Agent Harness

Give it a simple task and watch it operate your entire desktop. This project turns any AI into a GUI automation agent. You type a request like 'open the settings menu,' and it autonomously takes screenshots, spots buttons, and clicks for you. The cool part is its visual memory. It learns what interfaces look like after one look and remembers them later, so it gets faster every time. It even zooms in repeatedly to find tiny buttons you’d normally miss. It works on Mac, Windows, and Linux, using local or remote machines. It’s basically a digital assistant that can actually use a computer instead of just talking about it.

🆔 @hackernewsgithubprojects

📦 zero-mem/zero-mem

zero-mem: Peer-Reviewed AI Memory

Zero-Mem holds the official source code for a new memory system designed for large language models. Right now, the repository is quiet, showing only a promise that the full implementation and technical details will arrive after a rigorous peer review process. It does not offer code to download yet, nor does it explain the specific architecture or tricks it uses. Instead, it serves as a placeholder for researchers and developers waiting for a scientifically validated solution to model memory challenges. You can watch this space for updates, as the team is preparing to share their validated methods once the review is complete.

🆔 @hackernewsgithubprojects

📦 kirodotdev/kirocrew

Kiro Crew: The workspace that never forgets

Kiro Crew is the persistent workspace that never forgets a step. It acts as a memory for your development projects, storing task history and preferences so you never lose context when you log off. Instead of starting from scratch every time you return, Kiro Crew picks up exactly where you left off, learning from past mistakes to get smarter with each attempt. You can let it handle routine jobs unattended or chat with it through your favorite messaging apps. It runs locally on your own machine, keeping your data private while working hard in the background.

📰 https://news.ycombinator.com/item?id=49172209

🆔 @hackernewsgithubprojects

📦 zqxwce/vphone-ws

vPhone Workstation

Run a virtual iPhone right on your Mac without touching the command line. This native macOS app gives you a simple window to manage your research machines. You can browse existing setups, create new ones with a step-by-step guide, and boot the device with a single click. The coolest part is that the actual iPhone screen pops up in its own separate window, so you get a real Apple-like experience right on your desktop. It handles all the heavy lifting in the background, letting you focus on testing or experimenting with different iOS versions easily. Just install it and start exploring virtual Apple devices with zero hassle.

📰 https://news.ycombinator.com/item?id=49179328

🆔 @hackernewsgithubprojects

📦 ziangwu-77/reba

ReBA: Smarter Load Balancing for AI

ReBA is the smart load balancing technique that finally keeps vision and language tasks in perfect sync for large AI models. Standard AI models often struggle when switching between images and text, causing uneven workloads that slow everything down. ReBA fixes this by treating images and text as separate entities, ensuring that each part of the system gets a fair share of the work. It looks at how different types of data travel through the network and adjusts the routing so that no single part gets overwhelmed. This makes complex vision-language models faster and more efficient without needing extra hardware.

🆔 @hackernewsgithubprojects

📦 zhemdi/cadena

Turn Any 3D Mesh Into Editable Code

CADENA reverses engineering by turning a static 3D mesh into editable code. Instead of guessing the whole shape at once, it builds the model step by step, adding one operation like a cut or hole and checking the result against the original image. This stepwise approach ensures the final design is fully parametric and clean, solving the hard problem of reconstructing precise mechanical parts from rough scans. It effectively bridges the gap between visual data and functional engineering files for creators.

🆔 @hackernewsgithubprojects

📦 xiaomi-research/spatio-lm

SpatioLM: Spatial Intelligence for Vision-Language Models

SpatioLM lets AI models truly understand physical space in images and videos. It takes existing vision-language models and adds a lightweight module that teaches them about depth and 3D structure without needing special hardware. By learning from a teacher model that understands geometry, SpatioLM helps the AI answer questions like which object is closer or how things are arranged in a room. It works on standard photos and video clips, offering both reasoning skills and precise depth perception. This makes complex spatial tasks much easier for your favorite AI assistants to handle.

🆔 @hackernewsgithubprojects

📦 syhya/mlsys26-flashinfer-contest

How AI Agents Optimized CUDA Kernels for MLSys 2026

This is the MLSys 2026 FlashInfer Contest package that shows how AI coding agents can autonomously optimize complex computer chip instructions. It documents two winning approaches for generating high-speed graphics processing unit kernels. One method involves a human guiding the AI with specific skills, while the other lets an agent system work completely alone, planning and executing code without any human interference. The results are impressive, landing in the top three across multiple competition categories. You can see the exact step-by-step logs of how the AI solved these challenging coding tasks, offering a fascinating glimpse into the future of automated software engineering.

🆔 @hackernewsgithubprojects

📦 klboke/kkrepo

kkRepo: Your Self-Hosted Nexus Replacement

kkRepo lets you ditch the annoying limits of Sonatype Nexus by hosting every type of software package in one place. It supports Maven, npm, Python, Docker, Rust, Go, Helm, and many more, acting as a direct, compatible replacement for your existing setup. The real magic is the one-click migration tool that moves your users, permissions, and packages over without breaking a sweat, so your CI pipelines keep working exactly as they did before. It runs fast with low memory and handles large-scale storage via S3 or local files. If you are tired of component caps or slow upgrades, this is the clean, open-source fix you have been looking for.

📰 https://news.ycombinator.com/item?id=49179927

🆔 @hackernewsgithubprojects

📦 zyc-hust/csnet

CSNet Multi-Focus Image Fusion

CSNet is the machine learning tool that finally makes blurry photos sharp without guessing. It tackles multi-focus image fusion by intelligently picking the clearest parts of several shots and blending them into one crisp picture. Think of it as a digital camera that knows exactly which details are in focus, even when the lighting is tricky. The project lets developers run this process using standard Python and PyTorch, calculating key quality scores to prove the images look better. While some heavy calculations still rely on MATLAB, the core engine is ready to go.

🆔 @hackernewsgithubprojects

📦 halcyonnouveau/soppo

Soppo: Go With Better Safety

Write Go code that catches nil errors and missing enum cases before they crash your app. Soppo is a language that translates your code into standard Go, but adds powerful safety features Go lacks. You get automatic nil safety checks that stop pointer crashes at compile time, and pattern matching that ensures you handle every possible outcome. It also brings clean error handling and immutable data structures to keep your logic predictable. Since it compiles directly to Go, you can still use any existing Go library without changing your workflow. It’s basically Go with the safety net you wish it had from the start.

📰 https://news.ycombinator.com/item?id=49180168

🆔 @hackernewsgithubprojects

📦 nimi-research/tycho

Tycho: AI That Learns by Building World Models

Solve complex puzzles by building its own mental models of the game world. Tycho is an AI agent for the ARC AGI challenge that refuses to guess. Instead of just looking at pixels, it watches how things move, then writes a small Python script to describe exactly how the universe works. It tests this new understanding against what it sees, checking if its own invented rules hold true. When the model works, Tycho uses it to plan moves and solve problems it has never seen before.

🆔 @hackernewsgithubprojects

📦 jgraph/drawio-mcp

drawio-mcp

Artificial intelligence models used to struggle with visual design, but this project changes that by letting language models actually build diagrams. Draw.io MCP Server connects your favorite AI assistants directly to the popular draw.io editor. Instead of just describing a flowchart or system architecture in text, the AI generates the actual diagram files for you. You can create interactive visuals right inside your chat or open them in the full editor for editing. It supports complex shapes from cloud providers and standard diagrams alike. This makes explaining technical concepts or documenting software much easier. Your diagrams stay local, keeping your data secure.

📰 https://news.ycombinator.com/item?id=49180521

🆔 @hackernewsgithubprojects

📦 sandai-org/magi-2-preview

Magi-2 Preview: Video With Sound

Create ten-second videos complete with synchronized sound just by typing a description or uploading a single still image. Magi-2 Preview acts like a digital film studio that uses a massive neural network to dream up scenes while simultaneously composing the audio track you hear. It handles everything from text to video or image to video, stitching the picture and the noise into one polished clip. The real magic is that it generates the audio alongside the visuals, so you get a full sensory experience without needing separate tools. It is a clever step toward making content creation faster and more intuitive for everyone.

🆔 @hackernewsgithubprojects

📦 copilotkit/channels-sdk

Channels SDK

CopilotKit Channels SDK turns any chat app into a conversation with buttons and charts. You write the logic once and it works in Slack, Teams, Discord, or Telegram. The agent shows native buttons, lets you click, and pauses for your approval before acting. Your code stays yours while they handle the platform connection. It is a simple way to bring AI into your daily chat without leaving your usual apps.

🆔 @hackernewsgithubprojects

📦 ryanzhou/deepseek-v4-flash-mi300x

DeepSeek V4 Flash MI300X: The Patch That Runs In Production

DeepSeek V4 Flash MI300X is not a new software build but a precise collection of production patches that fix specific performance bottlenecks in the vLLM inference engine for AMD graphics cards. The repository contains only a readme and a folder of python files that act as direct replacements for internal components. These files are byte-for-byte identical to what actually runs in real servers today. The project solves the problem of getting large language models to run efficiently on specialized hardware without waiting for upstream updates. Instead of rewriting everything, the developer surgically tweaks specific mathematical operations and memory handling routines to squeeze out more speed.

📰 https://news.ycombinator.com/item?id=49166386

🆔 @hackernewsgithubprojects

📦 faramesh/faramesh-core

Stop AI Agents From Messing Up: Faramesh-Core

Faramesh-core lets you write simple rules that stop AI agents from making risky moves. Instead of trusting the model to behave, a local daemon checks every single action first. You define who can do what, like blocking account cancellations or limiting email sends. It enforces these limits strictly without needing cloud services. You get a tamper-proof log of every decision, so you know exactly what the agent did and why. It works with most agent tools you already use. This gives you real control over automated workflows, ensuring safety and clear accountability from day one.

📰 https://news.ycombinator.com/item?id=48023859

🆔 @hackernewsgithubprojects

📦 weavemindai/weft

Weft: The AI Programming Language

Weft is the programming language that finally lets AI systems pause for human input without writing a single line of webhook code. Think of it as a way to build complex software where artificial intelligence, databases, and actual people are all just equal ingredients. Instead of struggling to manage state or poll for answers, you simply drop a human query node into your workflow. The program stops, sends a request to a user, and then wakes up exactly where it left off days later, ready to continue. It handles the waiting automatically, so you can focus on the logic rather than the plumbing.

🆔 @hackernewsgithubprojects