📦 yosefhayim/ebay-mcp

Make eBay Answer to Your AI

Talk to your eBay store as naturally as you chat with a friend, letting your AI assistant handle the busywork directly. This local server connects your favorite coding assistant to eBay’s selling tools, handling the complicated security logins and data formatting automatically so you don’t have to. It transforms your account into a conversational interface, meaning you can ask to check inventory, update prices, or review recent feedback using plain language. You stay in control of what the assistant can do, keeping your business data safe while you focus on growing your shop.

🆔 @hackernewsgithubprojects

📦 zouyuxuan122/dsh-our-free-model

Our Free Model: Get Frontier AI Without Limits

Get access to top-tier AI models without typing a single login or API key. That is exactly what the Our Free Model plugin does for the dsh harness. It plugs in and gives you unlimited, free access to frontier models like Muse Spark and MiMo V2.6. The coolest part is that it actually measures your connection to show you which models will work right now, so you never hit a dead end. It even lets you force specific thinking budgets for deeper answers. No setup, no costs, just instant power. This is the easiest way to experiment with serious AI today.

🆔 @hackernewsgithubprojects

📦 pbs05/ponwrt

PonWrt: Unlocking Your Fiber Optic Modem

Take your locked fiber optic modem and turn it into a fully controlled router. PonWrt is a custom operating system built for specific hardware chips used in your internet box. It works where standard software fails, giving you direct access to the internal settings. You can stop the company from hiding network features and manage your connection exactly how you want. This project saves your original device data so you do not lose your internet connection after the upgrade. It turns a restricted box into a flexible tool for anyone who wants more control.

🆔 @hackernewsgithubprojects

📦 badtheorylabs/interference-search

Interference Search

Interference Search stops language models from getting lost in their own writing by treating every possible step as a live branch that moves forward together. Instead of guessing one path and hoping for the best, this method merges duplicate options and instantly cancels out dead ends before they waste energy. It turns a messy, winding trail of thoughts into a clean, synchronized wave that only keeps what actually works. The result is a reasoning process that solves hard puzzles in just three steps, while a standard model takes thousands of tokens to drift and fail.

🆔 @hackernewsgithubprojects

📦 scottrbk/forgetful

Forgetful: The Brain Your AI Agent Needs

Give your AI agent a brain that actually remembers things. Forgetful is an open-source tool that gives coding agents and chat bots a shared, long-term memory so they stop repeating the same mistakes or forgetting decisions you made hours ago. Instead of just storing raw text, it works like a digital notebook for your AI. Every time you save a note, the system automatically connects it to similar ideas it has already seen, building a web of knowledge that grows smarter over time.

🆔 @hackernewsgithubprojects

📦 anthropics/sandbox-runtime

Sandbox Runtime: Secure AI Agents Without Containers

Lock down a program so it cannot touch your secret files or connect to unauthorized websites, all without spinning up a heavy container. This is the core idea behind Anthropic Sandbox Runtime, a lightweight tool that wraps any command with strict operating system level boundaries. It is designed to make AI agents and local tools safer by default, meaning a script can only see and access exactly what you explicitly permit. The surprising part is how it handles the internet. Instead of just blocking signals, it routes every single network connection through a local proxy that checks each destination against your allowlist before letting it pass.

🆔 @hackernewsgithubprojects

📦 noelo-lab/kuna

Kuna: The Self-Improving Decompiler

Kuna is the decompiler built to be improved by AI. Most reverse-engineering tools are finished products, but this Rust-based tool was designed from the ground up so that other AI agents can actively refine its logic. Instead of a human expert manually tweaking every edge case, the project uses a built-in loop where AI agents test, find errors, and patch the code automatically. This turns a complex static analysis tool into a collaborative project where machines teach the software how to get better at reading machine code.

🆔 @hackernewsgithubprojects

📦 einsia/agent-git

Agent Git: Version Control for AI Chats

Stop losing your best AI coding conversations. Agent Git treats your chat history with coding assistants like code, giving it true version control. Instead of letting your hard-won insights vanish when a session resets, this tool saves every step, lets you rewind to earlier moments, and even lets you share a specific chat with a teammate. It works with popular assistants like Claude and Codex, turning messy, temporary text files into organized, resumable projects. You can pick up exactly where you left off or fork a conversation to try a different approach. It’s the difference between a fleeting note and a permanent record.

🆔 @hackernewsgithubprojects

📦 php-testo/testo

Testo: The PHP Testing Framework

Take full control of your PHP testing environment with Testo, a framework built on a lightweight core and a flexible middleware system. You get to keep the familiar, plain PHP syntax you already know while swapping out specific parts of the workflow to fit your exact needs. It handles tricky scenarios like retrying flaky tests or expecting specific errors through simple attributes, so you never have to fight your tools. If you are migrating from other suites, it even provides bridges to help you move over smoothly. The name comes from dough, symbolizing the malleability that lets you shape the testing process exactly how you want.

🆔 @hackernewsgithubprojects

📦 orailnoor/nativos

NativOS

Transform your rooted Android phone into a genuine Ubuntu Linux computer without flashing a new system. NativOS runs a complete touch-first desktop environment directly on your phone, letting you install real Linux apps alongside your normal Android apps. It handles the tricky hardware bridging automatically so you can get a full keyboard, file manager, and terminal access right on your screen. While it is still in alpha, the result is surprisingly capable for a device that was never meant to run a full desktop OS. If you have ever wondered how close a phone can get to a real laptop, this project shows you exactly where that line is.

🆔 @hackernewsgithubprojects

📦 menloresearch/asimov-1

Asimov 1

A fully open-source humanoid robot you can build yourself just landed, and it is surprisingly tangible. Asimov one is a thirty-five kilogram, human-sized biped with twenty-five moving joints, and the entire design is handed to you. You get the mechanical blueprints, the electrical wiring diagrams, and even a simulation file so you can practice making it walk before you touch the real thing. This is not just a kit; it is a transparent blueprint for a machine that can squat, lift, and turn its head. Whether you are a maker or just curious, the barrier to entry has never been lower for this kind of hardware.

🆔 @hackernewsgithubprojects

📦 amirmushichge/unreal-home-wizard

Snap a Photo, Walk Your Home: Unreal Home Wizard

Unreal Home Wizard is the photo-based tool that turns your living room into a walkable 3D world. Forget the steep learning curve of game engines; this wizard handles the heavy technical lifting for you. You simply feed it regular photos of your space, and it builds a realistic, navigable version of your home right in Unreal Engine. It is designed for total beginners, asking only simple questions while it sets up the complex software in the background. The coolest part is that you get to actually walk through your digital house, checking if the layout feels right before you finalize it.

🆔 @hackernewsgithubprojects

📦 feitangyuan/onetake

Onetake

Most AI video tools make product demos feel like loose slideshows, relying on lazy fades and constant cuts between scenes. Onetake is an ingenious Claude agent skill that fixes this by designing the exact boundaries where scenes meet, ensuring something on screen physically survives and morphs into the next beat. A prompt bar expands into an app window, a card grows into a page, and a single camera sweeps, shakes, and holds it all together in one unbroken shot. It even uses an automated oracle to measure continuity and reject videos that feel disjointed. It is a brilliant way to turn static screenshots into fluid, continuous masterclasses.

🆔 @hackernewsgithubprojects

📦 toon-format/toon

Token Oriented Object Notation

Pack massive amounts of data into your AI prompts without burning through your token budget. A lightweight data format called Token-Oriented Object Notation, or TOON, formats complex structures so they are incredibly compact and easy for artificial intelligence to read. Instead of wasting space on repetitive JSON keys or bulky XML tags, it condenses lists and nested objects into a clean, human-readable shorthand. Benchmarks show it slashes token usage by over forty percent compared to standard JSON while keeping model comprehension and retrieval accuracy just as high. It is a brilliant way to build smarter, faster, and cheaper AI applications without losing structure.

🆔 @hackernewsgithubprojects

📦 ctnicholas/jev-workflow-builder

Jev Workflow Builder: Collaborate on AI Workflows in Real-Time

Jev Workflow Builder lets you design AI workflows visually with your team, all in real time. The biggest surprise here is seeing your teammates’ cursors and live updates as you drag nodes around, making it feel like a shared whiteboard for automation. It connects to Jev and AI models so you can test runs instantly without writing code. If you’ve ever tried to explain an AI flow to a colleague and ended up in a slide deck, this tool cuts the noise. You build, test, and iterate together, side by side, with zero lag.

🆔 @hackernewsgithubprojects

📦 wb2024/wbs-guide-to-navidrome

WBs Guide to Navidrome

You can now run a private, high-fidelity music streaming service that completely replaces Spotify without sacrificing your data privacy. The open source project called wbs-guide-to-navidrome is a complete, masterfully curated roadmap for building your own personal cloud server using Navidrome. It guides you through using smart tools to auto-tag your music files, clean up messy metadata, and even use local machine learning to predict and inject mood tags directly into your tracks. You will learn how to build dynamic playlists without touching code, stream secure high-resolution audio to your phone on the go, and even generate a beautiful annual summary of your listening statistics just like Spotify wrapped.

🆔 @hackernewsgithubprojects

📦 ashhart/tensorfold

TensorFold: Speed Up Local AI Up to Three Times Faster

Speculative decoding usually forces you to trade away exact, deterministic outputs for pure speed, but tensorfold completely solves this dilemma. This clever engine serves local large language models on Apple Silicon and NVIDIA graphics cards using custom written kernels that deliver up to three times faster decoding than standard engines while remaining byte identical to standard serial decoding. It matches the exact tokens your model would normally produce one by one, ensuring zero loss in quality. It sets up an OpenAI compatible endpoint so your existing coding agents and apps can connect instantly, and even handles smart prompt caching to skip redundant processing on long conversations.

🆔 @hackernewsgithubprojects

📦 notoriouslab/vault-curate

Vault Curate

Vault Curate turns your Obsidian vault into a local first second brain that uncovers hidden connections between your notes without sending your data to the cloud. By running a compact one hundred and ten megabyte model directly on your machine, it analyzes your notes to suggest relationships you never explicitly linked. You can visualize these connections as interactive purple edges on an Obsidian Canvas, and with a single click, convert them into permanent wikilinks. It even categorizes your notes into hot and cold tiers to resurface forgotten ideas, making it incredibly easy to rediscover old thoughts.

🆔 @hackernewsgithubprojects

📦 heyokay/halobattery

HaloBattery: See Wireless Mouse Battery Without Vendor Software

Halo Battery lets you see the exact charge level of your wireless mouse or headset right in your Windows taskbar without installing Razer Synapse, SteelSeries GG, or any other brand specific software. Instead of hunting for hidden settings or running heavy background apps, this lightweight tool reads the battery data directly from your device's connection. It even shows a breathing green ring when your gear is charging, so you know at a glance that everything is powering up. This solves the annoying problem of guessing if your mouse is about to die mid game. No more surprise disconnects, just simple, clear battery icons for every device you plug in.

🆔 @hackernewsgithubprojects

📦 dorkitude/webctl

webctl: Save Tokens

Filter search results before they reach your agent, saving massive amounts of context. webctl is a command line tool that queries multiple search engines, then uses a scoring system to remove irrelevant links and duplicate content. It only returns the most useful snippets, meaning your AI doesn't have to read through junk. This cuts token usage significantly while keeping answer quality high. It works with any search provider and even handles rate limits automatically. If you are running local models or large language agents, this tool keeps your context window clean and your costs low. It is a simple way to make web search smarter for your code and chatbots.

🆔 @hackernewsgithubprojects

📦 engeldlgado/toshllm

ToshLLM: AI for Intel Macs

ToshLLM is the native macOS app that finally makes local AI work on older Intel Macs with AMD graphics. While most tools ignore these machines or produce broken text, this software fixes the underlying drivers to deliver fast, accurate results without cloud fees. It wraps powerful AI engines in a simple chat interface that runs entirely on your hardware. If you own an Intel Mac, this is your ticket to private, high-speed language models without buying new hardware.

🆔 @hackernewsgithubprojects