📦 bojieli/ai-agent-book

ai-agent-book

An LLM-driven AI agent can outperform traditional reinforcement learning by up to four hundred times in sample efficiency. The open-source repository ai-agent-book delivers a complete, bilingual textbook and hands-on code library that breaks down exactly how to build these intelligent systems. Instead of treating AI as a black box, it guides you through the core formula of combining large language models with context and tools. You will explore everything from optimizing memory and preventing prompt injection to building custom coding assistants and orchestrating multi-agent systems. It is the perfect, practical field guide for any curious developer wanting to turn raw AI models into highly capable, active assistants.

🆔 @hackernewsgithubprojects

📦 sirallap/agentglass

Agentglass

This lightweight desktop app turns your local machine into a real-time flight control tower for every AI coding agent running across your projects. Instead of letting terminal-based agents write code in the dark, agentglass maps their live actions on a single dashboard. It reads local project logs and system telemetry to visualize exactly what your AI assistants are doing. You can watch active tool calls, trace file changes in real time, inspect live terminal outputs, and monitor token usage and real dollar costs as they accumulate. It is a local-first mission control that gives you immediate oversight of your automated workspace.

🆔 @hackernewsgithubprojects

📦 apify/agent-skills

agent-skills

agent-skills is the developer toolkit that gives your AI coding agents hands-on web scraping superpowers. By dropping these specialized instructions into editors like Cursor or Claude Code, your agent can instantly scrape major social media, search, and map platforms using dozens of pre-configured templates. It handles the heavy lifting of selecting the right scraper, setting up inputs, and delivering clean, structured spreadsheets. Beyond ready-made scrapers, it helps your agent write new serverless scraping code in JavaScript or Python, and even packages your existing scripts into runnable cloud tools. It is the perfect bridge for turning simple prompts into reliable, automated data pipelines.

🆔 @hackernewsgithubprojects

📦 k-dense-ai/claude-scientific-skills

claude-scientific-skills

Over one hundred and sixty thousand scientists are using a single open library to convert standard AI assistants into capable research partners. The project, called claude-scientific-skills, equips AI agents with over one hundred ready-to-use scientific skills and direct connections to dozens of major databases across biology, chemistry, and medicine. Instead of just chatting, an agent can instantly lookup drug candidates in PubChem, analyze genetic data, or parse complex clinical trial reports. It bridges the gap between raw language models and specialized lab tools. It is an incredibly practical way for developers and researchers to build intelligent assistants that actually understand the language of science.

🆔 @hackernewsgithubprojects

📦 egeorcun/lucida

Lucida Background Remover

Keep the trickiest details perfectly intact when removing image backgrounds, from semi-transparent glass and glowing light effects to camouflage and intricate line art. Lucida is a smart, open-source tool trained specifically to handle the complex edge cases where standard background removers completely fail. By focusing on difficult textures like soft drop shadows on text, illustrations, and transparent bottles, it cleanly separates your subject without turning glass into an opaque blob. It is the perfect tool for creators and developers who need high-fidelity cutouts of complex objects. Check out the project to give your images a cleaner, professional look.

🆔 @hackernewsgithubprojects

📦 francislabountyjr/stereo2spatial

Stereo2Spatial

Stereo2Spatial is the audio conversion tool that uses specialized flow-matching AI models to transform standard stereo music into full spatial audio. If you want to experience your music library in a completely new way, this project trains and runs intelligent models that take mono or stereo tracks and map them directly into multi-channel setups like five-point-one or immersive twelve-channel layouts. The coolest part is how it processes sound: instead of chopping audio up awkwardly, it uses a smart method to sweep through a song smoothly, keeping the mix perfectly consistent from start to finish. You can easily turn your favorite stereo tracks into rich, three-dimensional soundscapes.

🆔 @hackernewsgithubprojects

📦 stupside/castor

castor

castor is the terminal tool that finally lets you beam any web video straight to your TV without dealing with buggy screen mirroring or laggy casting protocols. If you point it at a web page, it spins up a quiet, headless browser behind the scenes, sniffs out the actual high-quality video stream, and sends it directly to your TV. It even uses smart local processing to transcode the video and generate and burn in subtitles on the fly. It is perfect because it bypasses all the annoying smart TV limitations, giving you a smooth, full-resolution playback using just a single command.

📰 https://news.ycombinator.com/item?id=48964015

🆔 @hackernewsgithubprojects

📦 amirfish1/claude-command-center

Claude Command Center

Keep track of all your active local AI coding sessions at once instead of letting them get lost in a sea of terminal tabs. The claude-command-center project acts as a unified local dashboard for your Mac that automatically attaches to your running AI agents, including Claude Code, Codex, and Cursor. Because it reads their live state directly from your disk, it stays completely in sync even if you jump into a terminal and run things by hand. You can spawn tasks in parallel, see exactly which session needs your attention, and let different agents exchange context with each other automatically. It is a fantastic open source lens for developers looking...

📰 https://news.ycombinator.com/item?id=48428005

🆔 @hackernewsgithubprojects

📦 pablothethinker/hermespace

Hermespace

Give your digital assistant a permanent sense of self by recording its life in an append-only archive that never forgets. Unlike temporary chat histories that eventually fade or prune away, this project lets an agent build a deepening model of its beliefs, landmarks, and goals across every single session, ensuring it actually outlives the conversation. It works as a lightweight companion room that fits right into the agent's workspace, organizing focus, ranking active skills, and tracking decisions. Check out this clever workbench to help your helper remember who it is, where it has been, and where it is going next.

🆔 @hackernewsgithubprojects

📦 0xsero/litter

Litter

You can now run agentic coding workflows and manage remote development servers directly from your phone. Litter is a native mobile client for both iOS and Android that connects to local or remote servers so you can write and run code on the go. While the app interfaces are built natively using Swift and Kotlin, the entire backend state engine, server communication, and session management are handled by a single shared core written in Rust. It even supports real-time voice sessions and generative UI, giving you a full-featured portable command center that lets you manage your development environment from anywhere.

🆔 @hackernewsgithubprojects

📦 agi4sci/sciforge

SciForge

SciForge is a desktop platform designed to run AI research agents that can directly operate scientific software and analyze experimental data. If you have ever wanted an AI assistant that does not just talk about science but actually helps you do it, this is how we get there. It coordinates specialized workers for tasks like molecular viewing, bioimaging, sequence analysis, and scientific plotting. Instead of just querying a chatbot, you get an agent that can interact with complex structures, run computational workflows, and help manage physical lab data. It is a fascinating glimpse into how software will soon assist in discovery.

🆔 @hackernewsgithubprojects

📦 high-logic/genie-tts

Genie TTS

Genie-tts is a lightweight speech synthesis engine that finally brings lightning-fast voice cloning to standard computer processors. By optimizing heavy AI voice models for efficient runtimes, it cuts initial latency down to just over one second on a standard CPU and slashes memory requirements to a fraction of the original size. It comes pre-packaged with immediate, high-quality character voices in English, Japanese, Chinese, and Korean, and it can convert raw PyTorch files into efficient, deployable models instantly. You can now build incredibly responsive voice-activated software, interactive game characters, or real-time reading assistants that run fully locally on everyday hardware.

🆔 @hackernewsgithubprojects

📦 mtlab/pe-field

pe-field

pe-field is the open-source upgrade that takes traditional flat images and expands them into a fully structured three-dimensional space. While standard AI models struggle to understand depth when generating new angles of an object, this tool works by extending positional encodings into a volumetric field. It uses depth-aware and hierarchical structures to help diffusion transformers naturally reason about actual geometry. This means you can feed it a single flat picture and generate realistic new viewpoints or perform precise, spatially controlled image edits. It is a massive step forward for anyone trying to bridge the gap between flat 2D graphics and true 3D spatial awareness.

🆔 @hackernewsgithubprojects

📦 actava-ai/cura

Cura

Cura is a specialized one-trillion-parameter healthcare model that actually outperforms frontier systems like GPT-5.5 on demanding clinical benchmarks. Created through recursive self-improvement where AI training agents plan target capabilities and refine data based on observed failures, it excels at patient consultations, clinical reasoning, and interactive diagnoses. It even handles complex electronic health record tasks natively. What makes this project truly interesting is that the developers released the entire evaluation harness as an open-source, model-agnostic tool. This means you can plug in any OpenAI-compatible API or run local models with Ollama to run identical medical tests.

🆔 @hackernewsgithubprojects

📦 pixelcookie-zyf/taac-2026-serankmixer

taac-2026-serankmixer

Predict what users will actually buy or engage with after clicking an ad by using a highly optimized single-model ranking network. Built for the Tencent recommendation challenge, this project bypasses the need for massive, complex model ensembles to achieve a top-ten finish. It processes user behavior sequences and dense features using a specialized backbone combined with weighted-pair pooling to model data relationships more accurately. By implementing a dual-stream architecture with advanced training techniques like weight averaging and sharpness minimization, it delivers outstanding accuracy. It is a powerful, self-contained blueprint for building lightweight yet high-performing recommendation systems.

🆔 @hackernewsgithubprojects

📦 martindelophy/ai-video-editor

ai-video-editor

Edit full videos, generate AI voiceovers, and auto-transcribe subtitles entirely inside your web browser without uploading a single file to a cloud server. This clever tool brings a complete multi-track editor straight to your browser, keeping your media completely private. The coolest part is that it runs powerful machine learning models locally on your machine, letting you separate music tracks from vocals, crop videos automatically, and even generate talking digital avatars using your hardware. It is the perfect way to quickly slice clips, add natural-sounding voiceovers, and export your finished work securely without expensive subscription fees.

📰 https://news.ycombinator.com/item?id=48956143

🆔 @hackernewsgithubprojects

📦 qq456cvb/unipose9d

unipose9d

Track the exact 3D position, rotation, and actual size of absolutely any object in a photo without needing a pre-made 3D model or a category label. Normally, robots and cameras struggle to understand physical objects they have never seen before unless you feed them exact dimensions first. The unipose9d project changes that by estimating an object's full 9D pose from just a standard image and a simple point or text prompt. It figures out depth and alignment on the fly, making spatial tracking incredibly flexible. Grab the code and start tracking random objects instantly.

🆔 @hackernewsgithubprojects

📦 sagar-tamang/ultron-by-sagar-builds

ultron-by-sagar-builds

The browser interface that lets you rotate and zoom an Iron Man-inspired virtual orb using your webcam and hand gestures. Built on a Next.js and Three.js foundation, it uses modern artificial intelligence to track your hand joints on video, calculating the distance between your fingertips in real time. Pinching with one hand lets you grab and spin the beautiful orange wireframe globe, while squeezing both hands together or spreading them apart scales the view. It translates direct physical coordinates into smooth post-processed animations without complex accessories, giving you your own sci-fi setup right at home.

🆔 @hackernewsgithubprojects

📦 metareflection/formal-disco

Formal Disco

AI models can now automatically write, test, and self-correct mathematically proven software entirely on their own. While standard AI assistants frequently write buggy code, formal-disco uses a collaborative, multi-agent loop to generate complex programs that are guaranteed to work using rigorous verifiers like Dafny and Verus. It starts with an initiator agent that drafts a program, an automated verifier that tests it, and a specialized fixer agent that uses error feedback to patch up any flaws. By saving successful attempts, the system continuously fine-tunes itself to get smarter. It is a brilliant way to build ultra-reliable software datasets automatically.

🆔 @hackernewsgithubprojects

📦 no8d/comfyui-no8d-controls

ComfyUI NO8D Controls

The custom node pack comfyui-no8d-controls completely reshapes how you build workflows by packing complex sampler controls, image previews, and mask-based inpainting into a single, highly compact node. Instead of cluttering your workspace with a web of connected wires, it unifies your generation process with intuitive tools like an interactive brush, lasso, and eraser that trigger inpainting automatically when a mask is detected. It solves the headache of messy ComfyUI setups by streamlining LoRA stacking, visual style selection with responsive previews, and smart A/B image comparisons. It is the ultimate tool for keeping your canvas clean and your generation workflow fast.

🆔 @hackernewsgithubprojects