📦 vmdeshpande/ai-agent-automation

AI Agent Automation

AI Agent Automation lets you build AI workflows you can actually see and replay, not just trust blindly. It gives you a visual builder where you drag steps together, like sending an email or calling a model, and every single run is logged with inputs and outputs so you know exactly what happened. Because it is self-hosted, your data stays on your own servers, which is huge for privacy. If you have ever wondered why an AI did what it did, this is your answer. It turns unpredictable chatbots into transparent, inspectable tasks you can debug and rerun, giving you real control over the automation.

🆔 @hackernewsgithubprojects

📦 zaidmukaddam/fx-ui

fx-ui: The Desktop Home for Your Coding Agent

fx-ui is the desktop app that finally gives your coding agent a real workspace instead of a terminal window. It runs locally on your Mac, letting you talk to AI models like Grok or Codex through a clean interface where you can manage files, run code, and see every change your AI makes in a split view. The most surprising part is how it handles images, bypassing standard limits to let the model actually look at visual files. It is built for developers who want their AI assistant to feel like a proper software tool, not just a chatbot.

🆔 @hackernewsgithubprojects

📦 mac-automl/omniscope

Omniscope: Video AI Without the Lag

Processing long videos with AI models is usually a computational nightmare because the sheer volume of data chokes the system. Omniscope solves this by cleverly compressing the video and audio signals separately, which means your AI can understand entire scenes without bogging down. This approach is a big deal because it lets massive models handle real-time media without sacrificing accuracy. It is a brilliant way to make intelligent video processing actually feasible for everyday developers. If you want to build smart applications that watch and listen at once, this is the tool that makes it possible. Check out the repository to see how it works in action.

🆔 @hackernewsgithubprojects

📦 zostaff/agent-arena

DEGEN VILLAGE: The Trading Game

Build a trading bot in a cozy isometric village where you tune strategies in a forge and watch them trade with virtual funds. DEGEN VILLAGE is a free browser game that turns trading parameters into visual upgrades, letting you experiment with virtual money without risking a cent. The coolest part is the fly swarm, where four animated flies sit at monitors, their wings and legs moving in sync with live market data. It is a playful way to learn how trading works, proving you can build a strategy that actually makes money on paper. It is a clever, low-stakes sandbox for anyone curious about how markets move.

🆔 @hackernewsgithubprojects

📦 theseus-labs-rsi/awesome-rsi

Awesome RSI Maps AI Self Improvement

A single repository now organizes five hundred and nineteen academic papers into a clear hierarchy of how artificial intelligence systems can improve themselves. This curated collection solves the chaos of fragmented research by mapping every study to a specific level of autonomy, ranging from simple task adjustments to changing the very mechanism of learning itself. It is interesting because it moves beyond vague hype to define exactly what self-improvement means in practice. The project includes a companion survey that frames this capability as a roadmap, helping researchers distinguish between minor optimizations and genuine, persistent changes to an AI’s core architecture.

📰 https://news.ycombinator.com/item?id=49693565

🆔 @hackernewsgithubprojects

📦 yuvrajsingh-mist/smolcluster

Smolcluster: Train AI on Your Desk

Train and run large language models directly on the hardware already sitting on your desk, turning a mix of Mac Minis and Raspberry Pis into a single, powerful computer. Smolcluster solves the problem of needing expensive, dedicated data centers by using simple socket connections to split the heavy lifting of neural network training across whatever devices you own. The coolest part is how it finds and links these machines automatically, meaning no messy manual IP configuration or complex setup steps are required. You get a live view of your cluster and can launch different training algorithms without changing your code.

🆔 @hackernewsgithubprojects

📦 markfulton/ai-employees

AI Employees: Your Open Source Digital Staff

AI Employees is a set of plain text files that turn your existing AI agent into a full-time business worker on your own computer. This is the wild part: it does not just chat with you; it actually drives your web browser to fill out forms, draft emails, and manage your website, exactly like a human would. The most surprising feature is that these digital staff get smarter at your specific business every single time they run. If a webpage changes, the AI updates its own instructions so it keeps working without breaking. You do not need to hire a new tool or learn a new language.

📰 https://news.ycombinator.com/item?id=49785402

🆔 @hackernewsgithubprojects

📦 jakubantalik/thinking-orbs

Thinking Orbs: Stop Using Generic Spinners for Your AI

Stop using generic spinners when your AI is actually doing something specific. Thinking Orbs replaces that vague circle with nine distinct, hand-tuned animations that tell you exactly what your agent is doing. Is it searching? Solving? Weaving data? Each state has its own unique motion, rendered on a standard two-dimensional canvas so it runs smoothly everywhere without heavy graphics processing. It even automatically switches between light and dark modes to match your app, keeping the visual experience consistent. You get a clear, meaningful status indicator instead of a generic loading bar, which makes your interface feel much more responsive and thoughtful.

🆔 @hackernewsgithubprojects

📦 tnt-likely/panwatch

PanWatch: Your Personal AI Stock Analyst

Run a private stock market analyst on your own computer with PanWatch. It solves the problem of constantly checking prices by letting a team of specialized AI agents watch your portfolio for you. The most surprising part is the deep dive feature: when you click an icon, a simulated investment bank takes over. One agent studies the charts, another reads the news, and they literally argue about whether to buy or sell. After a few minutes of debate, a final manager makes the call and sends the reasoning directly to your phone.

🆔 @hackernewsgithubprojects

📦 estebanstifli/localtext2voice

Turn Books into AI Films with LocalText2Voice

LocalText2Voice is a free desktop app that takes a raw script and turns it into a finished, narrated video, all on your own computer. It solves the problem of expensive, cloud-only tools by letting you run powerful voice models locally, so your data never leaves your machine. The coolest part is the Video Storyboard feature: it doesn't just read the text, it actually analyzes the script to find characters and locations, then generates matching visuals and assembles them into a real movie. It’s like having a private film studio in your pocket, where you control every frame and every word.

🆔 @hackernewsgithubprojects

📦 aminmgmt/backpack

Backpack: The Tunnel That Hides in Plain Sight

Backpack connects your server to the outside world, but it does something clever: it can disguise itself as a normal website. If a scanner or a network filter looks at it, they just see a standard Nginx page, completely unaware there is a high-speed tunnel underneath. It is like hiding a secret radio channel inside a standard TV broadcast. Written in Go for performance, it uses encryption to make the data look like random noise or standard web traffic, so it slips past strict internet filters that usually block these connections.

🆔 @hackernewsgithubprojects

📦 kydlikebtc/awesome-grokbot

Awesome Grokbot: 730 Verified AI Automations

The entire ecosystem of shared Grok Bot automations is now consolidated into one verified catalog called Awesome Grokbot. Instead of digging through scattered social media posts, you get seven hundred and thirty live bots that are automatically checked daily to confirm they still work. Each entry includes a clear summary in both English and Chinese, so anyone can understand exactly what each tool does. It solves the problem of broken links by immediately flagging dead bots and moving them to a separate archive. You get a clean, searchable list of real automation ideas without the noise. Stop guessing which bots are still active and start building with confidence.

🆔 @hackernewsgithubprojects

📦 withlang-dev/open-jai

OpenJai: The Self-Hosting Language Compiler

This project is building a compiler that eventually writes its own source code in the language it compiles. OpenJai is a new systems language designed for speed and developer control, but right now it is just a bootstrap tool written in Zig. It takes your code, breaks it down into steps, and hands it to LLVM to turn into a runnable program. The crazy part? The team is writing the actual compiler logic in the new language itself, right next to the old one. It is a race to see if the new compiler can build itself without help. If it works, you get a fully independent toolchain from scratch.

📰 https://news.ycombinator.com/item?id=48116458

🆔 @hackernewsgithubprojects

📦 cseti007/comfyui-crossviewwarp

CrossView Warp: The 3D Camera Orbit for AI Video

Take a standard video and physically move the camera around the subject without re-shooting. CrossView Warp is a ComfyUI node that acts as a visual depth-warp engine, letting you drag a marker on a 3D sphere to choose a new angle. It reprojects your clip into that new view, filling in the gaps where the original lens never looked. The coolest part is the live preview: you orbit the scene in real time, seeing exactly what the AI will generate before you even start. It turns a flat recording into a dynamic, multi-angle shot, solving the stiff, locked-camera problem that plagues current AI video tools.

🆔 @hackernewsgithubprojects

📦 kingmaniya/comfyui-llm-text-processor

LLM Text Processor for ComfyUI

Feed images straight into local language models inside ComfyUI without leaving your node graph. This plugin, called the LLM Text Processor, lets you drop in GGUF models like Qwen or Gemma to automatically write prompts, translate text, or generate image captions. The real kicker is that it handles multimodal tasks right on your own hardware. You can plug in a batch of photos and let the AI look at them, all while running completely offline. It even separates the model’s internal reasoning from its final answer, so you see exactly how it thought. If you are building automated image workflows, this gives you a powerful, private brain for every step.

🆔 @hackernewsgithubprojects

📦 yifanzhang-pro/klpo

KLPO: The No-Frills AI Trainer

KLPO is the reinforcement learning method that cuts the noise out of AI training. Most algorithms need a whole group of answers to figure out what is right, but this one gets by with just a single attempt. It solves a real problem where you need to teach an AI to handle tasks, but you do not want to waste time calculating the value of every step. It works by using simple math to estimate the score without needing a separate judge model. This is fascinating because it makes complex training much more direct and easier to run on your own computer.

🆔 @hackernewsgithubprojects

📦 pacesalab/bindcraft2

BC2: The Tool That Designs Proteins From Scratch

BC2 designs entire protein binders from scratch, creating new molecules that lock onto specific biological targets without needing a pre-existing template. It solves the problem of finding the right molecular key for a lock by generating and testing thousands of candidate designs in one go. You simply describe your target and pick a binder style, like a small antibody or a cyclic peptide, and it handles the complex math of folding and binding. This is useful because it turns a months-long trial-and-error process into a structured workflow that spits out ranked, testable sequences. Instead of guessing what works, you get a list of predicted complexes ready for the lab.

🆔 @hackernewsgithubprojects

📦 saganaki22/comfyui-hyperflow

Generate Video and Sound in Eight Steps

A single plugin lets you generate video and synchronized audio using only eight distinct steps, a massive speedup for standard video models. This tool, called ComfyUI Hyperflow, adapts a specific training method to work directly inside the ComfyUI environment without modifying its core files. It achieves this by adding a unique timing condition that guides how the model integrates data at every single stage. This approach preserves the distinct relationship between time and the generation process, ensuring the output matches the original high-quality results. The result is a faster, integrated workflow for creators who want to produce both moving pictures and matching sound without needing complex external setups.

🆔 @hackernewsgithubprojects

📦 ghuntley/preflight

Preflight: Stop Leaking Secrets to AI

Preflight is the security guard that stops your API keys from accidentally leaking to AI models. It sits quietly between your applications and the cloud, scanning every request and attached file for sensitive data like passwords before the model ever sees them. If it finds a secret, it blocks the transmission or redacts the sensitive text on the spot. This simple check prevents catastrophic data breaches caused by copy-pasting code or uploading documents that contain hidden credentials, making it an essential safety net for anyone building with large language models.

📰 https://news.ycombinator.com/item?id=49776416

🆔 @hackernewsgithubprojects

📦 teorth/optimizationproblems

Optimization Problems: Tracking the World's Hardest Math Numbers

Optimization Problems is the collaborative math tracker that finally pinpoints the exact edges of our knowledge in pure mathematics. The repository catalogues best-known upper and lower bounds for dozens of notoriously difficult constants, turning scattered academic papers into a single, living ledger. This is useful because it lets researchers and curious minds see precisely where the frontier lies, from the moving sofa constant to the matrix multiplication exponent, and it openly invites the crowd to submit improvements. When a new proof tightens a boundary, it updates the record, creating a transparent map of progress.

🆔 @hackernewsgithubprojects