📦 karminski/awesome-llm-benchmark-prompts

awesome-llm-benchmark-prompts

The repository awesome-llm-benchmark-prompts contains a highly creative suite of coding prompts specifically designed to test whether advanced artificial intelligence can handle ultra-complex visual effects, physics simulations, and 3D animations. Instead of boring text tests, these prompts challenge AI models to generate complete, single-file interactive web demonstrations of complex scenarios using tools like Three.js and Pygame. These challenges range from simulating the complete physical chain reaction of underwater firecracker explosions to rendering the orbital disintegration of the moon. It is a fantastic benchmark for developers who want to push the boundaries of what code-generation models can realistically build.

🆔 @hackernewsgithubprojects

📦 posecode-dev/posecode

Posecode

Turn plain text into smooth, anatomically correct 3D human animations right in your browser. This project introduces a lightweight language designed for physiotherapy, yoga, and fitness training that lets you describe physical exercises in simple text. Instead of relying on heavy, unpredictable AI video generators, you write clean, timed steps detailing joint angles and movements. The magic happens behind the scenes, where an intelligent rendering engine automatically clamps every joint to safe, realistic ranges of motion, preventing impossible bends. It even features an AI-powered critic to analyze your movement scripts for balance and biomechanical realism. It is the ultimate tool for developers looking to build interactive, lightweight physical guides.

🆔 @hackernewsgithubprojects

📦 datagallery-lab/enterprise_agent_platform

Enterprise Agent Platform

Move your AI prototypes into reliable production systems by studying a comprehensive architectural blueprint designed for real-world operations. The enterprise_agent_platform repository is a highly detailed, open-source book and companion codebase that guides developers through the complex reality of running artificial intelligence in business environments. Instead of simple chat windows, you will learn how to design state machines, manage permissions, control costs, handle unexpected system failures, and set up continuous evaluations. It covers everything from data pipelines to secure user interfaces, helping you transition from fragile demos to stable, enterprise-ready systems.

🆔 @hackernewsgithubprojects

📦 robinhood-ape/robinhood-noxa-bundler

robinhood-noxa-bundler

The multi-wallet launch bundler that beats public snipers to the punch, robinhood-noxa-bundler is built specifically for launching tokens on the Robinhood Chain. When you launch a token on a decentralized market, third-party sniper bots usually buy up the cheap supply instantly. This clever tool solves that by automating a coordinated, multi-wallet purchase. It funds your secondary wallets, launches your new token, and fires off simultaneous buy orders directly into the newly created liquidity pool. It coordinates everything in a flash so you actually secure your own supply before outside bots can react. Give it a spin on dry-run mode to see it coordinate the magic.

🆔 @hackernewsgithubprojects

📦 pattern-ai-labs/agentcall

AgentCall

Your AI coding agent can now join your video meetings and talk directly with your team. A clever tool called agentcall makes this possible by letting popular developer assistants step into Google Meet, Microsoft Teams, and Zoom as full participants. Instead of just reading text, your agent listens to real-time transcripts, speaks naturally with low-latency text-to-speech, and can even share its screen to present dashboards or code changes. It keeps its full workspace context, meaning it can run commands and edit files in the background while discussing the project live. It is a brilliant way to turn your agent into a collaborative team member.

🆔 @hackernewsgithubprojects

📦 pablostanley/yoinks

Yoinks Video Downloader

Grab high-quality videos from over eighteen hundred sites directly inside your terminal without dealing with shady ads or confusing download websites. This clever command-line tool bypasses the sketchy popups of the web to let you download video or audio files in seconds. You just paste a link from YouTube, Instagram, or TikTok, choose your preferred resolution, and it handles the rest. It runs a clean, interactive screen right in your command line and automatically bundles all the heavy-lifting tools it needs. Try it out to instantly build a clean, local media archive with zero hassle.

🆔 @hackernewsgithubprojects

📦 rajtilakjee/kivo

Kivo Desktop Teleprompter

You can now display a floating, translucent teleprompter overlay directly on your desktop that automatically updates whenever you edit your script file. This lightweight python application, called kivo, solves the frustrating problem of losing your place during video recordings or presentations by providing a sleek, frameless window that stays on top of your other apps. It features smooth auto-scrolling with easily adjustable speeds and quick keyboard shortcuts to pause or resume. It is perfect for reading scripts or presentation notes naturally without cluttering your screen, leaving you with a distraction-free way to deliver flawless presentations every single time.

📰 https://news.ycombinator.com/item?id=48714810

🆔 @hackernewsgithubprojects

📦 antins-labs/searchos

SearchOS

AI search agents easily get stuck in infinite, brain-melting loops when they fail to find a piece of evidence. That is why searchos operates like a literal operating system for your web-searching bots. Instead of letting individual agents wander blindly, this framework externalizes the search progress into a shared, persistent state. It tracks tasks, visualizes gaps on a coverage map, and remembers past failures so your bots never repeat the same useless queries. By scheduling and overlapping tasks like CPU processes, it systematically fills in data gaps without wasting your API budget. It is basically a smart manager that keeps AI searchers organized, efficient, and actually helpful.

🆔 @hackernewsgithubprojects

📦 structureclaw/structureclaw

StructureClaw

The AI-assisted structural engineering workspace that converts plain language into computable building designs, structureclaw makes modeling physical structures incredibly intuitive. Imagine describing a two-story steel frame over chat and watching the system automatically generate, validate, and analyze the virtual 3D model. It bridges the gap between conversational AI and rigorous engineering by handing tasks directly to trusted engines like OpenSees or commercial analysis tools. It is a brilliant way to quickly draft models, run static checks, and export detailed calculation reports without starting from scratch in complex software. You should definitely check out this repository to see how AI is reshaping physical design.

🆔 @hackernewsgithubprojects

📦 jinyangwu/seed

seed

Artificial intelligence models can now analyze their own past mistakes in real time to train themselves without human guidance. A new framework called seed allows a single AI model to play two roles simultaneously during training: it interacts with its environment to complete tasks, and then analyzes its own performance to extract hindsight skills. These self-discovered skills are used to immediately update the model's decision-making logic through a joint reinforcement learning process. Because the analyzer update happens dynamically, the feedback stays aligned with the model's changing behaviors, creating a highly efficient self-evolving system that requires absolutely no complex prompt engineering or external databases during deployment.

🆔 @hackernewsgithubprojects

📦 alibaba/atrex-kernel-agent

Atrex Kernel Agent

Writing high-performance GPU code is incredibly difficult, but atrex-kernel-agent makes an artificial intelligence agent do the heavy lifting for you. This clever project sets up an automated loop that takes basic PyTorch code and iteratively tunes it into a screaming-fast GPU kernel. Instead of guessing where the lag is, the agent runs real hardware profiling, consults a massive built-in database of GPU specs, drafts a targeted optimization plan, and safely tests changes until it hits peak performance. It is like having a world-class hardware engineer automatically tuning your code while you drink coffee. You get maximum speed with none of the usual headaches.

🆔 @hackernewsgithubprojects

📦 gnomeria/usbtree

USB Tree

Monitor your USB devices right from your terminal without ever needing root permissions or heavy background drivers. This lightweight tool displays a gorgeous, live-updating visual tree of your hardware, showing you exactly how hubs and devices are chained together across Linux, macOS, and Windows. You get color-coded speed badges, instant visual alerts when devices are plugged or unplugged, and even safe-eject capabilities. Linux users can even watch live data traffic sparklines for each port in real time. It is a brilliant, zero-dependency utility that makes debugging hardware actually fun. Grab it and inspect your workspace today.

🆔 @hackernewsgithubprojects

📦 selfimproving-agent/awesome-self-improving-agents

Awesome Self-Improving Agents

Artificial intelligence can now rewrite its own code, memory, and prompts to get smarter over time without human intervention. The awesome-self-improving-agents repository is a curated map of research papers and projects dedicated to systems that learn from their own experiences. Instead of relying on static programming, these systems use feedback loops to optimize their tools and prompts, and even run simulated environments to practice tasks before executing them. It is an incredible resource for developers who want to build software that actively teaches itself. Check it out and see how the next generation of autonomous systems is training itself.

🆔 @hackernewsgithubprojects

📦 klingairesearch/metaview

MetaView

Generate realistic new viewpoints of an object or scene from a single photograph, even when making massive camera angle changes. While traditional methods struggle to keep shapes and scales consistent when you rotate the camera, metaview solves this by combining a smart geometry model with a clever spatial anchoring technique. It extracts natural depth clues from the image and uses them to guide a diffusion model, keeping the scene locked to a real physical scale. This prevents the object from distorting or drifting as you shift the perspective. It is perfect for anyone wanting highly realistic, mathematically consistent view rotations using just a single picture.

🆔 @hackernewsgithubprojects

📦 alibaba/atrex-bench

Atrex-Bench

Atrex-Bench is the evaluation benchmark that finally puts AI-generated GPU code to the test using real-world performance limits. Instead of just checking if an AI can write basic code, this tool takes actual production workloads and challenges AI models to compile them into hyper-optimized, low-level GPU kernels. It supports major backends like Triton and CuteDSL on both NVIDIA and AMD hardware. Crucially, it grades the AI's code against the absolute speed-of-light physical limits of your graphics card, showing you exactly how close the machine-written code gets to flawless hardware efficiency.

🆔 @hackernewsgithubprojects

📦 jia-ethan/claude-keysmith

claude-keysmith

The Claude Code helper that cleanly injects custom instruction files into your AI memory without breaking anything else. Instead of letting complex custom rules clutter up your projects, claude-keysmith acts as a tidy manager that places individual Markdown rules into their own isolated folder and gently inserts removable, marked import blocks into your CLAUDE.md files. It runs completely locally with built-in safety defaults like automatic back-ups and zero-touch dry runs. You get perfectly organized, modular instructions across user or project levels, and uninstalling a rule only removes its specific block, leaving the rest of your files pristine.

🆔 @hackernewsgithubprojects

📦 six-bit-tx/salientgs

SalientGS

Reconstruct highly detailed 3D scenes from unordered photos faster and more accurately using an elegant, budget-friendly pipeline. Traditionally, converting flat images into 3D environments requires slow preprocessing and struggles with blurry, underfit areas. SalientGS fixes this by introducing an smart, importance-guided algorithm that automatically detects poorly reconstructed regions and reallocates a fixed budget of virtual 3D pixels directly where they are needed most. By combining camera pose estimation and 3D rendering into one unified, GPU-accelerated step, it delivers stunningly realistic 3D scenes in just ten minutes. You get crisp, highly accurate results without needing massive computational power.

🆔 @hackernewsgithubprojects

📦 trail-ml/agent-cow-python

agent-cow-python

Intercept your AI agent's database writes and keep your production data completely safe. When you build autonomous agents, giving them direct access to your database is incredibly risky, but this clever library solves that by isolating all of their modifications in a temporary copy-on-write layer. Your AI agent can run queries, insert rows, and delete files exactly like normal, but nothing actually touches your real database until you manually review and approve the changes. You can easily roll back mistakes, cherry-pick the good edits, or commit everything at once, making agent integrations incredibly safe and predictable.

🆔 @hackernewsgithubprojects

📦 apple/ml-mmtoolsandbox

ml-mmtoolsandbox

AI assistants can now navigate complex visual interfaces just like humans, thanks to a benchmarking system developed by Apple. The project, called ml-mmtoolsandbox, lets developers evaluate how well visual AI agents can use over five hundred tools across sixteen different application domains. Instead of just reading text, these models have to look at images, process multi-turn conversations, and execute precise tool calls to solve tasks. It simulates realistic challenges like goal revisions and error corrections in real time. This tool makes it incredibly easy to see if an AI can actually manage a complex workflow without losing track of the visual details.

🆔 @hackernewsgithubprojects

📦 oversecured/samsung_vulnerabilities

samsung_vulnerabilities

Peek behind the scenes of mobile security to see how preinstalled apps can compromise a phone. The samsung_vulnerabilities repository details a collection of one hundred and seventy-six security flaws discovered in factory-installed apps on Samsung devices. From settings bypasses to silent file access, it showcases real bugs found and fixed during a multi-year security collaboration. It is a goldmine for understanding how tiny oversights in system privileges can leave a device open to exploits. Check it out to see exactly how these security flaws work under the hood and level up your own defensive coding skills.

🆔 @hackernewsgithubprojects

📦 xiyuzhai-husky-lang/visored

Visored

Write mathematical proofs in a natural, controlled language and let a computer automatically translate them into rock-solid code. That is what visored does by bridging the gap between human-readable math and formal logic. It takes intuitive mathematical arguments and transpiles them into incredibly detailed, machine-verified Lean code. Because the translation is so thorough, a short, forty-line human proof can expand into tens of thousands of lines of rigorous, step-by-step logic. It is basically a super-powered translator that makes formal math verification accessible without requiring you to write thousands of lines of manual code yourself.

🆔 @hackernewsgithubprojects