168 subscribers
11K videos
11.6K links
Download Telegram
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ elpapi42/pi-observational-memory

Pi Observational Memory: Endless Sessions

An AI coding agent can stay sharp for weeks without forgetting a single decision. Pi Observational Memory fixes the silent amnesia that happens when long coding sessions hit their context limit. Instead of summarizing your entire history in one slow, lossy burst, it quietly watches your work in real time, picking out key decisions and turning them into durable facts before you even need them. This means when the system finally has to compress your session, it already has a clean, accurate snapshot ready to go, so the handoff is instant. You stop losing the thread of your logic after a few hours of debugging.

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

OpenClaw: The AI Agent That Actually Works

OpenClaw is the self-hosted Docker image that gives your AI agent a real, persistent body. It solves the frustration of running complex AI tools in containers by baking in a complete, automated setup that survives restarts. You get a full browser environment with a visible desktop, allowing the AI to log into sites, handle two-factor authentication, and maintain active sessions just like you do. The entire system, from chat history to configuration, lives on a single persistent volume so your data never vanishes when the container is recreated.

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

Ascii Sharks: A Living ASCII Ocean

Sharks swim across your screen rendered entirely in text characters. This project, called Ascii Sharks, builds a living ocean simulation using only a code grid and zero external libraries. It features eight distinct species, from classic great whites to rare thresher sharks, each with unique shapes and movements. The simulation uses mathematical currents to guide the fish, creating a natural, flowing experience. It also respects accessibility settings, offering a static view for users who prefer reduced motion. This is a fascinating look at how creative coding can turn simple characters into a complex, animated world. The takeaway: even basic tools can create surprisingly rich digital art.

πŸ†” @hackernewsgithubprojects
Media is too big
VIEW IN TELEGRAM
πŸ“¦ k0valik/pi-blackhole

pi-blackhole: Free AI Memory

Compress long AI conversations without losing crucial details, thanks to pi-blackhole. This tool solves the frustrating problem where artificial intelligence forgets important decisions after summarizing a lengthy chat. Instead of asking a model to write a summary, which is slow and costly, this project uses a mathematical method to shrink the text instantly for free. While compressing, three background helpers quietly read the conversation to remember key facts. They save these notes so your agent stays smart and consistent, even after hours of work. It turns a chaotic, forgetful chat into a reliable, long-term memory system. Give your AI a true long-term memory by installing pi-blackhole today.

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

Reticle: The AI QA That Catches Lies

Your AI coding agent just told you the fix is done, but it has never actually opened your app to see if it works. Reticle solves this by giving your AI the ability to look at the live software, click buttons, and check the network traffic behind the scenes. Instead of just taking its word for it, the tool drives your real running application and returns a clear verdict: pass, fail, or cannot tell. This is crucial because it catches silent errors like a broken button that still loads the page but does nothing when clicked.

πŸ“° https://news.ycombinator.com/item?id=50163896

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ memtensor/memt-1.0

MemT-1.0: The AI That Actually Remembers

Give your AI a permanent memory that doesn't bloat your hard drive. MemT-1.0 is a new family of foundation models that keeps a constant-sized memory state no matter how long your conversation gets. Instead of storing every single word you ever typed, it compresses your chat history into a tiny, evolving snapshot. This means you can talk to it for hours, save that memory to a file, and pick up exactly where you left off later without reloading the entire transcript. It even remembers how to code by keeping past experiences in its head rather than pasting them into the prompt.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ pc2005-cloud/dsh-pet

dsh-pet

This tiny desktop pet lives directly on your computer screen and physically reacts when you flick it with your mouse. Developed as a plugin for a coding assistant tool, this project lets you install a transparent, animated character that wanders your desktop, sleeps, and even checks your API balance. It is built from the ground up so you can generate your own custom pets using AI video. The most interesting part is the simple setup that puts a living, interactive creature right on your screen. If you want a friendly companion for your workspace, this is exactly what you need.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ mihirgupta0900/tv-home

tv-home

Transform a sluggish, ad-cluttered Android TV into a snappy, Apple TV-style experience by remotely stripping out twenty-five background apps and installing a minimal home screen with rotating nature wallpapers. This project connects over Wi-Fi to disable manufacturer junk and telemetry without permanently uninstalling anything, keeping every change fully reversible. On a two-gigabyte device, this cut the home screen memory usage from one hundred ninety-two megabytes down to just eighty-three. It is the perfect way to reclaim speed and clarity for your living room display. Give your old TV a fresh, fast second life today.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ mihawk-99/ps5_mihawkos

Windows 11 on a PS5? Yes. MihawkOS

MihawkOS is the homebrew system that puts a full Windows desktop on your PlayStation five. Forget just playing games, this project gives your console a real operating system with a taskbar, start menu, and file explorer, all styled like Windows eleven. It uses a custom runtime to run actual Windows programs, meaning you can launch sixty four bit apps and even older thirty two bit games right from your console. The coolest part is that it handles graphics natively, so your DualSense controller and touchpad work just like a mouse and keyboard.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ panchovix/nvidia-windows-vbios-vix-loader

NVIDIA Windows VBIOS Vix Loader

NVIDIA Windows VBIOS Vix Loader is the registry trick that swaps your graphics card firmware without ever touching the physical chip. Normally, changing your GPU's power limits or clock speeds requires physically flashing the card's internal memory, which is risky and permanent. This tool instead tells the Windows driver to grab a modified version from your system settings, letting you test new settings safely and revert them instantly if something goes wrong. It works on modern cards like the RTX 50 series and even some older models. You can tweak power limits and check if the changes are actually active all from a simple window.

πŸ“° https://news.ycombinator.com/item?id=50157484

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

The Coolest Valorant Skin Browser

This is the Valorant Skin Changer, the web tool that finally organizes all those scattered weapon skins into one clean spot. It takes the messy hunt for Vandal and Operator looks and puts them in a simple catalog you can actually browse without clicking through twenty different tabs. The real magic is how it lets you compare cosmetic themes side by side, so you can finally see which bundle matches your vibe before you spend a single credit. It is not just a list, but a visual playground for your in-game style. If you want to stop guessing and start curating your collection, this is the place to go.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ isurumaduranga/harness-engineering-101

Harness Engineering 101

Build a working AI agent in just three hundred lines of plain Python. Harness Engineering 101 strips away the intimidating jargon to reveal the simple loop behind every intelligent assistant. You send a list of messages, the model replies, and your code updates that list. That is it. The model has no memory, so your program acts as the body that keeps the conversation alive. Each chapter adds one specific fix for a real problem, turning a basic chat box into a capable tool user. By the end, you will have built the core logic yourself and finally see that there is no magic, just careful data management.

πŸ“° https://news.ycombinator.com/item?id=50151390

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

Ondine

Compile raw Mermaid syntax directly into lightweight SVG vector graphics before sending a single byte to the browser. Instead of downloading heavy layout engines and client-side parsers that slow down web pages, this high-speed compiler handles all the diagram processing ahead of time during the build process. A tiny two-hundred-byte script mounts the finished vector diagrams into the document object model, keeping loading times fast while preserving full styling flexibility through custom CSS variables. It transforms complex technical charts into lean assets without compromising browser performance. Try integrating precompiled vector diagrams into your next site build for an instant speed boost.

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

frontmcp

frontmcp is the TypeScript framework that gives artificial intelligence model context protocol servers the enterprise structure they usually lack. Building tools for language models can feel like the wild west, forcing developers to hack together custom authentication, job queues, and cloud deployments from scratch. This tool brings full backend structure to model context protocol development with built-in dependency injection, input validation using Zod, and production security controls. The coolest part is how easily you can deploy your project anywhere from Node to serverless platforms like Cloudflare Workers and AWS Lambda without rewriting a single tool. Give it a try on your next AI app.

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

mjfit

mjfit is the physics simulator tool that finally builds exact assemblies for real-world robotics. When training AI to plug in a USB cable or mesh tiny mechanical gears, standard simulators usually chop hollow shapes into rough chunks that block the opening or lag down your computer. This clever framework cuts plates around holes into exact convex wedges, making round pegs and tight sockets fit together perfectly without fake collisions. It runs an order of magnitude faster than traditional mesh methods while staying ridiculously accurate down to micrometer tolerances. If you work with physics engines, this makes complex mechanical training smooth and reliable.

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

Torch TPU: Run PyTorch on Google's TPUs

Run your PyTorch models directly on Google's Tensor Processing Units, and Torch TPU makes it happen. This official backend bridges the gap between standard PyTorch code and TPU hardware, so you can train and infer on these specialized chips without rewriting your entire workflow. It handles the heavy lifting of compiling your model graph for optimal execution and supports distributed training across multiple devices. The best part is that it integrates with the standard compile function, letting you leverage existing tools while getting TPU speed. If you want to explore large language models on TPU infrastructure, this is your go-to starting point.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ dweep4269/sideraven

Sideraven: Vertical Tabs for Android

Sideraven transforms the standard Android tablet experience by bringing a vertical tab sidebar to your screen. This project is a fork of the Iceraven browser, designed specifically to give users a layout inspired by the desktop Zen Browser. Instead of hiding your open tabs in a cluttered menu, you get a clean, organized list running down the side of your display. This change makes multitasking far more intuitive, allowing you to see and switch between your work at a glance. It is a simple yet effective way to reclaim screen space and organize your digital life.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ zoran123456/lola-retro

Lola Retro: The True CRT Effect

Take retro games and make them look exactly as they did on a real television back in the nineties. Lola Retro is a shader for RetroArch that stops at just adding scan lines and instead recreates the entire analog signal path, from the console's encoder all the way through to the glass of the tube. The most surprising part is how it handles the subtle rainbow shimmer and color bleeding that happened when a game used a cheap composite cable.

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

ComfyUI on Your Phone?

Run professional AI art workflows right from your pocket with CueForge. This project takes the powerful ComfyUI engine and wraps it in a dedicated mobile interface, letting you tweak prompts, connect nodes, and generate images without hunching over a desktop keyboard. It solves the friction of needing a big screen to experiment, putting the full creative toolkit in your hand. You can watch generations process in real time and refine your setups instantly. It turns a complex desktop tool into something you can actually use on the go. If you create with AI, this is the bridge you have been missing.

πŸ†” @hackernewsgithubprojects
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ“¦ aiwithremy/claude-skills-llm-council

LLM Council: 5 AI Advisors for Better Decisions

LLM Council is the Claude Code skill that turns one question into five expert opinions. Instead of getting a single, potentially biased answer, this tool spins up five distinct AI advisors who each attack your problem from a completely different angle, like a skeptic, an optimist, or a pure logician. They don’t just guess; they anonymously peer review each other’s work to catch blind spots and push back on weak reasoning. Finally, a chairman synthesizes everything into a clear verdict, telling you exactly where the group agrees, where they clash, and the one concrete step you should take next.

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

Hello Halo: The AI That Actually Works On Your Computer

hello halo is the desktop agent that actually has hands to do real work on your computer. Most AI chatbots just talk back at you, but hello halo sits on your machine, keeps running all day, and can open files, browse the web, and even handle your emails or chat apps. It connects your phone so you can check in on what it’s doing from anywhere. The coolest part is that it doesn’t just answer questions; it takes action, like finding a file or filling out a form, right on your screen.

πŸ†” @hackernewsgithubprojects