📦 alikon-art/determinflow

DeterminFlow: Deterministic Workflows for Probabilistic AI

DeterminFlow is the workflow runtime that finally brings structure to unpredictable AI models. Most AI tools act like a single overworked assistant who forgets details and costs a fortune as the chat grows longer. DeterminFlow solves this by breaking complex tasks into isolated, versioned nodes where each step only sees the context it needs. This means you get automatic retries, precise cost tracking, and the ability to resume exactly where you left off if something fails. It is already saving developers massive amounts of tokens in real production pipelines. It turns chaotic AI experiments into reliable, shippable services that just work.

🆔 @hackernewsgithubprojects

📦 codegraphcontext/grapharc

GraphARC: The Safety Gate for AI Agents

GraphARC is the governance layer for AI agents that finally stops wild code execution before it happens. Most tools just watch AI agents make mistakes and hope for the best, but this project installs a strict admission gate between the plan and the run. An artificial intelligence proposes a workflow, and a deterministic checker immediately reviews it against your safety policies and budget. If a step violates your rules, the gate rejects it and refuses to execute, forcing the agent to rewrite its plan. You get a complete audit trail that matches your live dashboard exactly, so you know exactly what cost and what logic was used.

🆔 @hackernewsgithubprojects

📦 maxteabag/sqlit

Sqlit: The Terminal SQL Client That Feels Like Magic

Sqlit is the terminal app that finally makes database queries fun. It works like a lightweight version of LazyGit but for SQL. You just type sqlit and it pops up a clean, colorful screen right in your terminal. The coolest part is how it spots your running Docker containers automatically. You press Enter, and it connects without you needing to type complex server addresses or ports. It even remembers your passwords safely and lets you use familiar keyboard shortcuts to type code. It is fast, beautiful, and makes looking at data actually enjoyable instead of tedious.

🆔 @hackernewsgithubprojects

📦 0xwilliamortiz/humanizer-cli

Spot AI Text in Your Terminal

Humanizer-cli turns your terminal into a detective for AI writing. It checks your drafts against thirty-three habits that give away language models, like fake enthusiasm and endless em dashes. The tool runs entirely offline with zero dependencies, so your text never leaves your machine. It spots obvious mechanical tells instantly and highlights subtler patterns that need a human reader. You can scan a file, search for specific quirks, or get a ready-to-paste prompt to rewrite the text by hand. It is a clever, lightweight way to ensure your writing sounds like you, not a machine.

🆔 @hackernewsgithubprojects

📦 ash-2046/suv

SUV: Future Scene Understanding as Video Generation for End-to-End Driving

A single front camera can now predict a car's entire future path by generating four video streams simultaneously. The project called SUV treats driving planning as a video generation task. It looks at what is happening right now and renders a four-second preview of what the road will look like, including lane markings, depth, and moving objects, all in one go. Instead of guessing many possible paths and picking the best one, the system simply watches its own generated video to figure out where to steer. This approach avoids complex selection logic and achieves top scores on autonomous driving benchmarks.

🆔 @hackernewsgithubprojects

📦 shrek3onvh5/minimax-h3-nativeaudio-musicvideo-workflow

MiniMax H3 Native-Audio Music Video Workflow

MiniMax H3 Native-Audio Music Video Workflow is the ComfyUI toolkit that locks generated music videos to your original song files instead of letting AI hallucinate new vocals. While many video generators scramble the audio, this project adds a special node that forces the video’s lip movements to match the exact rhythm and timing of your supplied track, keeping the audio pristine throughout. It also includes a multishot sampler that chains scenes together so the character stays consistent, plus a clever speed boost that predicts video details between steps to render longer clips faster.

🆔 @hackernewsgithubprojects

📦 aut-aisl/queenvis

QueenVIS: Video Segmentation Without Video Training

QueenVIS teaches a computer to track objects across video frames without ever seeing a video. Instead of learning from moving footage, it trains exclusively on static images. It achieves this by adding two simple rules to its object detection queries: predicting where the object’s center is and matching its visual features to a single snapshot. This trick makes the model so confident about an object’s look and position that it can guess which object is which in a moving video later on. By removing these rules after training, it relies on a clever memory system to link frames together.

🆔 @hackernewsgithubprojects

📦 xfreeze2/quill

Quill Dictation for Mac

Quill lets you speak directly into any Mac app without ever touching your clipboard, placing your voice exactly where you click or at the end of existing text. It uses your existing Grok subscription so you skip extra API keys, and it even spots voice commands like opening a new coding session mid-sentence to keep your flow going. The tool is lightweight, self-contained, and keeps your privacy tight by storing nothing but local history. It is a surprisingly neat way to type faster while keeping your hands on the keyboard.

🆔 @hackernewsgithubprojects

📦 tanvirnwu/sureflow_iros_2026

SUREFlow: Tiny Robot Model, Giant Results

SUREFlow: Tiny Robot Model, Giant Results

Meet SUREFlow, a robot manipulation model that punches way above its weight class. While most robot brains the size of a small app have hundreds of millions of parameters, this one only has about one hundred seventy-nine million. It is built on a clever architecture called Mamba, which lets it process visual information and language instructions much faster than traditional neural networks. But here is the real kicker: it does not just work well on practice tasks; it actually understands what to do when things change.

🆔 @hackernewsgithubprojects

📦 brightdata/cli

Bright Data CLI: Terminal Web Scraping

Bright Data CLI lets you scrape any website, run searches, and extract structured data directly from your terminal without writing code. This official command-line tool handles tough anti-bot protections, CAPTchas, and JavaScript rendering automatically, so you get clean markdown or JSON instead of messy HTML. It also includes powerful features like AI-driven scraper creation and self-healing fixes when data breaks. You can even integrate it with coding assistants via MCP for smarter workflows. With a free monthly tier that covers most basic tasks, it is an accessible way to pull reliable web data for personal projects or research.

🆔 @hackernewsgithubprojects

📦 kenn-io/msgvault

msgvault: Your Personal Message Archive

msgvault lets you archive a lifetime of email, chat, and calendar history into one local, offline database. Instead of letting decades of messages get lost behind web interfaces, it pulls everything into a searchable archive powered by SQLite and DuckDB. You can run millisecond analytics, search your entire history, or even ask AI questions using your own local models. It is the only way to truly own your correspondence.

🆔 @hackernewsgithubprojects

📦 thorodinson246/whatisit-nl2sh

Whatisit-nl2sh: Local AI Shell Command Generator

Whatisit-nl2sh is the local AI tool that turns plain English into shell commands without sending data to the cloud. Instead of searching through manuals or guessing syntax, you simply type a request like find files bigger than 100MB and it generates the exact command instantly. It runs entirely on your machine using a small model that fits in less than one gigabyte, meaning your typing history never leaves your computer. This makes it perfect for privacy-conscious developers who work in restricted networks or just want a fast, offline way to automate terminal tasks.

🆔 @hackernewsgithubprojects

📦 mbeher2200/darkhours

DarkHours: Your Astrophotography Planner

DarkHours is the astronomy planner that finally tells you exactly when and where to shoot the night sky. Most apps just show a moon phase, but this one does the heavy math on sky brightness, light pollution, and cloud cover to give you a single quality score for any night. It simulates the real sky with thousands of stars and shows you the best shooting windows for deep sky objects. It even finds darker driving spots nearby if your current view is too bright. There is a 360 degree sky view you can scrub through time, plus a red mode that protects your night vision.

🆔 @hackernewsgithubprojects

📦 dondai44423/bladebro

Bladebro: AI Browser Driver with Re-Render Immunity

Bladebro lets artificial intelligence agents browse the web without getting caught or confused when pages update. Most bots lose their grip the moment a website reloads its interface, forcing a restart. Bladebro fixes this by assigning every clickable item a permanent structural fingerprint, so the AI recognizes buttons even after they vanish and reappear. It uses just five simple commands to navigate, read, and click, returning only what changed to save computing power. Built in Rust, it runs as one tiny file with zero extra software. It quietly mimics human typing and mouse movements to slip past detection systems.

🆔 @hackernewsgithubprojects

📦 lifecycle-innovations-limited/claude-ops

Your Business, Run by AI

claude-ops is a business operating system for Claude Code that turns your developer environment into a fully autonomous startup. It runs twenty-one specialized agents handling everything from merging code to tracking revenue. The most striking feature is the unified inbox. It pulls messages from WhatsApp, email, Slack, and Telegram into one stream. Your AI agents read these messages and act on them immediately. You get a marketing optimizer, a revenue tracker, and an infrastructure monitor working together without lifting a finger. It effectively lets you run a whole company with a single voice assistant. This isn't just a chatbot.

🆔 @hackernewsgithubprojects

📦 rfandango/xendroid

XenDroid Android Xbox 360 Emulator

XenDroid is the Android emulator that finally brings Xbox 360 gaming to your phone with a custom Kotlin interface that makes the complex simple. Most emulators feel like wrestling with a technical manual, but this project wraps the powerful Xenia Edge engine in a sleek, touch-friendly app you can actually use. The standout feature is its smart per-game configuration system that automatically tweaks settings to fix crashes and missing graphics for specific titles without ruining performance for others. You can just drop your game files in, tap to play, and let the app handle the heavy lifting.

📰 https://news.ycombinator.com/item?id=49243649

🆔 @hackernewsgithubprojects

📦 wtfox/luna.nvim

luna.nvim

Luna does something surprisingly quiet for a developer tool: it replaces the busy, colorful clutter in your code editor with a deep, near-black background that lets you focus entirely on what you are writing. Instead of overwhelming you with bright syntax highlighting, it uses a sophisticated grey scale with just a tiny, intentional splash of warm and cool colors for things like keywords and functions, making the text pop without distracting you.

📰 https://news.ycombinator.com/item?id=49238486

🆔 @hackernewsgithubprojects

📦 bigmacfive/turbo-graph

Turbo-Graph: Constrained Search for Smart AI Memory

Turbo-Graph is the AI search tool that finally handles complex filters without slowing down your app. Most retrieval systems force you to fetch a huge list of data first, then throw it all away to match your specific rules. Turbo-Graph changes that by searching inside your tags, dates, and sources all at once. It builds a smart index of your data and remembers which combinations you have already checked, so it does not waste time on the same questions twice. If you are building an AI assistant that needs to be accurate and fast, this is the upgrade your memory layer has been waiting for.

🆔 @hackernewsgithubprojects

📦 os-factory/har

HAR: The Open Harness for Multi-Agent Coding Workflows

HAR is the open-source harness that lets multiple coding agents work on the same project at the same time without stepping on each other's toes. Most tools struggle when you try to run several AI assistants together because they fight over shared databases, ports, or git files. HAR solves this by giving each agent its own completely isolated workspace, or worktree, so they never collide. It uses a simple contract file to tell agents exactly how to build and test the code, then runs a strict verification check to prove the work is solid.

📰 https://news.ycombinator.com/item?id=49210031

🆔 @hackernewsgithubprojects

📦 antigmalabs/ante

Ante: The Lightweight Coding Ghost in Your Terminal

Get coding in your terminal without the heavy baggage. Ante is a tiny, self-contained agent that fits in just fifteen megabytes and runs entirely on your own machine. Unlike popular alternatives that rely on massive cloud dependencies, Ante bundles its own inference engine right into a single Rust binary. This means you can run complex coding tasks fully offline using local models, with zero API keys and no internet connection required. It is shockingly light on resources, using a fraction of the memory and CPU that competitors need. It feels like having a ghost in your shell that works hard without slowing you down.

🆔 @hackernewsgithubprojects