📦 mengto/skills

Skills

Skills is the prompt-engineering playbook that finally helps your AI agents build gorgeous web interfaces and playable games without losing their minds. Instead of relying on vague instructions and hoping for the best, this repository packages hyper-specific design systems, animations, and gameplay logic into reusable markdown files. Your AI helper can read these exact blueprints to handle complex tasks like scroll-driven storytelling, custom particle effects, or even responsive three-d web graphics. It basically gives your favorite coding assistant a masterclass in high-end design, ensuring you get polished, production-ready code on the first try.

🆔 @hackernewsgithubprojects

📦 jgravelle/jcodemunch-mcp

jcodemunch-mcp

jcodemunch-mcp is the model context protocol server that slashes artificial intelligence token costs by over ninety-five percent during code exploration. Instead of feeding entire files into your artificial intelligence assistant, this clever tool uses tree-sitter syntax trees to pull only the precise code symbols, imports, and relationships you actually need. It plugs directly into tools like Claude Code and Cursor, meaning your assistant can map out project architecture, trace dependencies, and find dead code without burning through your token budget. It keeps your context window clean and your development workflow incredibly fast.

🆔 @hackernewsgithubprojects

📦 different-ai/openwork

OpenWork

Share your AI workflows, skills, and connected services with your teammates instantly using a single tool. OpenWork acts as an open-source control center for your AI agents, allowing you to bundle custom skills, database connections, and API access into a single Model Context Protocol connection. This means you can build a workflow once and run it across different platforms like Claude Code, Cursor, or ChatGPT without rebuilding anything. It solves the headache of fragmented AI setups by giving your team a unified marketplace for custom tools. Grab the code, hook up your favorite model, and elevate your team's collective intelligence today.

🆔 @hackernewsgithubprojects

📦 max-sixty/worktrunk

Worktrunk: Git Worktrees Made Easy for AI Agents

Ever tried running five AI agents at once only to have them trip over each other's code? The tool called worktrunk solves this by completely rethinking how we use git worktrees. Instead of clunky native commands where you have to type paths and branches three times over, this rust-based tool lets you spin up, switch, and tear down isolated environments with simple branch names. It even copies ignored build caches like node modules or target folders so your agents do not suffer through slow cold starts. It is the perfect control center for parallel workflows, keeping your main directory pristine while your digital helpers build in parallel.

🆔 @hackernewsgithubprojects

📦 evolution-foundation/evolution-go

Evolution Go

Evolution Go is the high-performance messaging engine that brings lightweight, blazing-fast WhatsApp integration directly to your backend. Built from the ground up in Go, it completely bypasses resource-heavy setups to deliver minimal memory usage and rapid response times. The standout feature is its native support for passkey pairing, allowing you to link accounts locked behind modern biometric security using a simple browser helper. It easily handles real-time events through webhooks or message queues, manages media storage via compatible cloud buckets, and provides clean API endpoints for sending interactive carousels, lists, and media status updates. Experience lightning-fast, production-ready WhatsApp connectivity today.

🆔 @hackernewsgithubprojects

📦 vibepod/vibepod-cli

VibePod CLI

VibePod CLI is the terminal tool that finally makes running AI coding agents completely hassle-free. Instead of cluttering your local system with global packages or worrying about security, this utility launches popular coding assistants like Claude, Gemini, or Copilot inside isolated containers using Docker or Podman. It unifies all your favorite helper agents under a single command, automatically tracking their HTTP traffic and usage metrics. Everything gets saved directly to a local dashboard, giving you a clear, privacy-first view of what your agents are doing. It is the perfect setup for easily comparing different models on your real-world projects safely.

📰 https://news.ycombinator.com/item?id=48894261

🆔 @hackernewsgithubprojects

📦 taucetiproject/tauceti

Tau Ceti

Imagine a world where human mathematicians write the high-level roadmaps while artificial intelligence does the heavy lifting of writing and verifying the actual code. That is the reality behind Tau Ceti, an open-source mathematical library built on the Lean theorem prover. In this unique ecosystem, humans set the mathematical goals and design strict review rubrics, while AI agents write the proofs, submit code, and even conduct adversarial reviews against each other to find errors or weak arguments. It is a fascinating experiment in cooperative, machine-driven science that ensures math is verified at an incredible scale while remaining entirely free for everyone to build upon.

📰 https://news.ycombinator.com/item?id=49047064

🆔 @hackernewsgithubprojects

📦 cerebrasresearch/reap

reap MoE Compression

Researchers can now cut the size of massive mixture of experts language models in half while maintaining near-lossless performance on complex tasks. While traditional compression techniques merge neural network experts together and cause massive drop-offs in accuracy, a new tool called reap shows that intelligent pruning is actually the superior path forward. By analyzing both router gate values and average expert activation norms, reap calculates exactly which experts contribute the least to a model's output and safely removes them. On massive neural networks, this elegant approach successfully trims fifty percent of the experts with virtually no drop-off in coding or reasoning skills.

🆔 @hackernewsgithubprojects

📦 cloudflare/nimbus

Nimbus

Build beautiful documentation sites where humans and AI agents are both treated as first-class citizens. Instead of hiding your site files inside a complex dependency, this project scaffolds a complete Astro site directly into your repository, giving you full ownership over every layout, style, and component from your very first commit. It automatically generates search-engine-friendly pages alongside agent-readable formats like LLM text files and structured data, ensuring coding assistants and AI web crawlers can navigate and understand your documentation just as easily as a human reader. You get total control to edit anything you want, backed by a powerful framework that handles the heavy lifting behind the scenes.

🆔 @hackernewsgithubprojects

📦 encod3d-sec/claudebrain

Claudebrain

Claudebrain is the interactive hacking companion that turns your personal knowledge base into an AI-powered security assistant. By combining your markdown notes in Obsidian with custom AI skills, it acts as a smart engine for penetration testers and bug bounty hunters. Instead of getting lost in endless checklists, you can ask your AI to automatically draft bug disclosures, suggest payloads, or analyze security logs using your own curated guides. It organizes complex testing methods, active directory tricks, and web attack paths into a highly searchable brain. It is like having a seasoned security partner who remembers every single exploit technique and helps you plan your next move.

🆔 @hackernewsgithubprojects

📦 thecodacus/understory

Understory

Build a self-wiring markdown memory that lets your local AI agents store, link, and organize what they learn as they work. Understory functions like a living knowledge graph beneath your models, storing every single fact as a human-readable markdown file with clean yaml formatting. Instead of dumping raw data into a basic vector database, it actively cross-links new concepts, flags broken links, and automatically weaves orphan notes back into the main structure. You can browse your memory using a gorgeous interactive force-directed graph or replay exactly how your agent navigated through its notes during a task. It runs flawlessly on local models, connects to Claude as an MCP server, and...

🆔 @hackernewsgithubprojects

📦 arman-bd/chromiumfish

Chromiumfish Stealth Browser

Most open source stealth browsers get caught by detectors because they try to hide using easily spotted JavaScript patches. Chromiumfish is different because it moves all of its fingerprint spoofing directly down into the C++ engine of Chromium itself, leaving absolutely no tampering marks for bot checks to find. It presenting a completely coherent browser identity tied to a single persona seed. Best of all, it integrates directly with Python and Node using your existing Playwright scripts, and it even comes with a natively built AI agent that can solve plain-language tasks and bypass tough bot checks on its own. Give your scrapers a genuine cloaking device.

📰 https://news.ycombinator.com/item?id=48462366

🆔 @hackernewsgithubprojects

📦 misaka-mikoto-tech/agent-skills

agent skills

If you are building AI agents to browse the web or run local commands, you need to check out agent skills. This collection gives your AI assistants superpowers, like a robust tool to read Bilibili video pages by extracting official subtitles, danmaku stats, and comments. If a video lacks subtitles, it automatically downloads the audio stream using Node modules and transcribes it locally. It also includes a bulletproof PowerShell invocation template that teaches your Windows agents to safely handle native arguments, spaces, and encoding without breaking. It is a fantastic starting point for building reliable, self-sufficient AI agents.

🆔 @hackernewsgithubprojects

📦 achrefelouafi/waterthreejs

Water ThreeJS

Build a stunning, photorealistic ocean simulation directly in the browser without loading a single texture. This project generates everything procedurally using advanced shaders, including crashing waves, realistic sky reflections, and drifting volumetric clouds. By translating complex physics into mathematical formulas, it dynamically calculates light absorption so shallows look bright turquoise while deep waters fade into a rich, dark blue. When you dive below the surface, the simulation seamlessly transitions to show sunlit god rays, rolling caustics on the sandy seabed, and floating objects that naturally bob and drift on the water's surface.

🆔 @hackernewsgithubprojects

📦 mighty-alien/mev-arbitrage-bot

MEV Arbitrage Bot

The mev-arbitrage-bot is an automated trading tool that scans Ethereum decentralized exchanges to profit from temporary price differences. By combining a smart contract with an external Python script, the tool monitors live Uniswap events for trading opportunities. When a price mismatch occurs between different trading pools, the system can buy low in one venue and sell high in another within a single transaction. Users deploy the contract using a web interface, fund it with ether, and run the automated scanner. It serves as a practical, hands-on example of how modern decentralized finance searchers automate on-chain trading.

🆔 @hackernewsgithubprojects

📦 davidhdev/canvas-ui

Canvas UI

You can now run realistic fluid simulations and glass distortion effects directly over your live, interactive web page. An open-source library called canvas-ui makes this possible by utilizing a new web browser feature that lets canvas graphics directly capture and redraw your active website. Instead of dealing with unresponsive background videos, your text remains completely selectable and your links stay fully clickable while being distorted in real time. It comes packed with dozens of custom creative effects, and it works across all major modern frameworks. It is a fantastic way to add some serious visual polish to your next project.

🆔 @hackernewsgithubprojects

📦 baidu-baige/loongflow

LoongFlow AI Agent Framework

Build expert-grade AI systems that actually learn from their own mistakes instead of just repeating them. LoongFlow organizes artificial intelligence into a structured planning, execution, and summary loop, meaning the agent constantly reflects on what went wrong and how to fix it. It even uses specialized memory to store and reuse these hard-won insights across complex coding, machine learning, and math tasks. Think of it as giving your local AI model the rigorous, reflective mindset of a human scientist. You can use it to build incredibly stable, self-improving agents that tackle your toughest problems with genuine, evolving logic.

📰 https://news.ycombinator.com/item?id=49049823

🆔 @hackernewsgithubprojects

📦 chinmay-sawant/gopdfsuit

gopdfsuit

Full compliance archival PDF generation delivers an astonishing sixty-six hundred operations per second on a single machine, while turning off compliance reaches an unbelievable thirty-seven thousand operations per second. This incredible speed comes from gopdfsuit, an open-source engine that works as a fast Go library, a REST API, or Python bindings. It generates fully tagged, digitally signed, and accessible documents directly from JSON templates, completely in-memory with zero external runtime dependencies. It is the perfect modern toolkit for any developer needing to quickly generate, merge, secure, or redact highly compliant documents at scale. Check out the gopdfsuit repository to speed up your document workflows today.

🆔 @hackernewsgithubprojects

📦 jbwinters/jacquard-lang

Jacquard Lang

Run and verify untrusted code written by machine learning models without risking your system. This project introduces a small programming language called jacquard-lang that is built specifically for a world where AI writes the code and humans review it. Instead of guessing what a model-generated script might do, this language explicitly lists every outside-world effect, like network or file access, directly in the function signature. If an effect isn't declared, the code won't compile, and if it isn't granted at runtime, it won't run. It is the ultimate safety net for reviewing AI code with absolute peace of mind.

📰 https://news.ycombinator.com/item?id=48894630

🆔 @hackernewsgithubprojects

📦 naver-ai/opd2

On-Policy Delta Distillation

Teach smaller artificial intelligence models how to reason by capturing only the specific changes that occur during a larger model's training. This approach uses a technique called on-policy delta distillation, which isolates the exact reasoning shift between a smart teacher model and its basic version. Instead of copying everything, the system focuses solely on this difference, or delta signal. Tests show this method consistently boosts performance in math, science, and coding tasks across various model sizes. It offers a much smarter, more direct way to pass down complex thinking skills to lightweight models without the usual extra bloat.

🆔 @hackernewsgithubprojects