📦 ontos-ai/knowhere

Knowhere: The Memory for AI

Knowhere is the document engine that turns messy files into clear memory for AI. It reads your PDFs and slides, organizing them into a smart map instead of just random text. This means your AI assistant doesn't just guess what is on a page. It actually knows where things are, like a librarian with perfect recall. This is huge because it stops AI from hallucinating facts about your documents. It makes the data reliable and easy to trace back to the source. If you are building agents that need to read real papers or reports, this is the tool that makes them smart. Stop feeding raw files to your models.

📰 https://news.ycombinator.com/item?id=49738595

🆔 @hackernewsgithubprojects

📦 gcwing/openbitfun

OpenBitFun: Your AI That Never Stops

Keep your AI working across your laptop, phone, and even your smartwatch without losing your place. OpenBitFun solves the annoying problem of switching devices by letting you start a task on your desktop and finish it on your phone, keeping the context alive. It is not just a chatbot; it is a full workspace where you can build custom tools directly for your daily needs. The coolest part? You can actually ask the agent to write new code to extend the app itself. If you want an AI partner that adapts to your workflow and never loses its train of thought, this is the tool to try.

🆔 @hackernewsgithubprojects

📦 jakeschincariol/linkedin-agent-skill

Linkedin agent skill

Write LinkedIn content that actually sounds like a human. This open source pack gives Claude eleven specific instructions to help you draft posts, comments, and weekly plans. The real stand out is the humanizer tool. It scans your draft to remove invisible characters and overused words that scream artificial intelligence. It then scores your text on five different checks to make sure it reads naturally before you hit send. You stay in full control, deciding exactly what goes out. If you want to sound like a person and not a bot, grab this repo and try the humanizer on your next post.

🆔 @hackernewsgithubprojects

📦 typesafeainate/dspy-typesafeify

dspy-typesafeify: One Decorator, Smarter AI

dspy-typesafeify is the Python decorator that makes your AI programs think faster by routing specific answers to a specialized logic engine instead of relying entirely on the main language model. By simply adding one line of code to your existing setup, it identifies which parts of a task need strict, reliable logic and handles them separately, while still letting the AI generate free-form text where it excels. This hybrid approach cuts processing time and costs significantly without forcing you to rewrite your application.

🆔 @hackernewsgithubprojects

📦 maurodruwel/nimstats

NIMStats: The AI Benchmark That Runs Itself

NIMStats is the free benchmark dashboard that automatically tracks real-world performance for twenty-two major AI models. It solves the problem of trusting marketing claims versus seeing actual speed and reliability data by running a simple test on every model every single hour. The coolest part is that it requires zero infrastructure because it uses free automated tasks to do all the heavy lifting and hosts the results on a static site. You get a live leaderboard showing who is fastest and who is most reliable without paying for any servers.

🆔 @hackernewsgithubprojects

📦 honda-research-institute/hrdx

How Honda's HRDX Teaches Cars to See Maps from the Sky

Honda’s research team just dropped HRDX, a massive new benchmark that helps self-driving cars build their own maps using only their cameras. The standout feature? They prove that looking down from a plane helps. By training with aerial imagery, the system learns to see the road better without needing any extra sensors on the car itself. It’s like studying the answer key to ace the test, then walking away without the cheat sheet. This dataset covers fourteen hundred kilometers of real driving, making it a huge leap forward for safe autonomous navigation.

🆔 @hackernewsgithubprojects

📦 shi3z/deepseekv4.1-a100-custom

DeepSeek V4.1 Flash: 1 Million Tokens on 4 GPUs

DeepSeek V4.1 Flash is squeezing a full million-token context onto just four A100 GPUs without crashing. This custom inference runtime is interesting because it proves you don't need a massive cluster to handle ultra-long documents on standard hardware. It works by chopping the prompt into small chunks and pipelining them across the cards, preventing the memory explosion that usually kills these runs. The result is a stable, offline server that processes massive contexts reliably. If you are building tools that need to analyze huge files, this project shows that high-end long-context performance is accessible right now.

🆔 @hackernewsgithubprojects

📦 immortalhowwl/gptheist

Ten Agents, One Veto: Inside GPTHEIST

A single market decision passes through ten named checks, and any one of them can stop the entire process. GPTHEIST is a read-only observation tool for tracking new token launches on Robinhood Chain. It simulates a strict approval chain where distinct roles verify data, liquidity, and social signals before a final coordinator renders a judgment. Crucially, the system cannot place real trades; it only records paper-only decisions. This design creates a transparent, deterministic audit trail for analyzing how multi-agent systems handle risk. It offers a clear blueprint for building safe, verifiable workflows without touching live funds, making it a valuable study in coordinated decision-making.

🆔 @hackernewsgithubprojects

📦 apple/xcode-project-format

Apple Replaces Xcode Project Files

Apple just gave developers a typed Swift library to read and write the new JSON based project files that replace the old messy plist format. Instead of guessing how Xcode stores data, this tool models every folder, file, and build setting as a precise object, so your scripts can safely inspect and change projects without breaking the build. It solves the problem of fragile text parsing by providing a shared, accurate model that generators and validators can rely on. You can finally trust your automation to handle Xcode projects without reverse engineering the file structure by hand.

📰 https://news.ycombinator.com/item?id=49735329

🆔 @hackernewsgithubprojects

📦 linguo2625469/workbuddy2api-panel

WorkBuddy2API Panel: Auto-Complete Tasks

WorkBuddy2API Panel turns Tencent accounts into an OpenAI-compatible gateway, but the real surprise is that it automatically completes seventeen out of eighteen official growth tasks for you. This self-hosted tool wraps your existing account into a standard API endpoint, letting any compatible client connect without changing a single line of code. It handles the heavy lifting in the background, like auto-logging in and managing multiple account pools so you never hit a rate limit. The most impressive part is a single button that clicks through all the tedious daily activities, from chat logs to expert calls, and claims the rewards for you.

🆔 @hackernewsgithubprojects

📦 lucidrains/rlt

RLT: The Looped Transformer

Loop your neural network layers to think longer without stacking more hardware. This project, called RLT, is a PyTorch implementation of the Recurrent Looped Transformer. It reuses the same set of layers multiple times, effectively creating a much deeper model from a smaller footprint. This is brilliant because it lets you build massive, reasoning-heavy models on standard consumer GPUs. You get the depth of a giant network with the cost of a tiny one. It is a clever way to squeeze more intelligence out of the same silicon. If you want to experiment with deep thinking models that actually fit on your machine, this is the one to check out.

🆔 @hackernewsgithubprojects

📦 yujunzhixue/dsh-purge

dsh-purge: The Tool That Unlocks Every Model

dsh-purge is the local harness plugin that finally lets every model speak freely by stripping away the refusal locks that often come from the host layer, not the model itself. Many users blame the AI for saying no, but the platform wrapping it might be the real blocker. This tool maps those guardrails and applies reversible patches to your local setup so you can see exactly where the limits are coming from. It works with different models and lets you swap prompts per model to test behaviors without breaking your install. It is not an attack kit, but a research aid for understanding how your local environment shapes the conversation.

🆔 @hackernewsgithubprojects

📦 kunkka19xx/lgtm

lgtm: The Code Review Tool Built for AI Agents

lgtm is the terminal reviewer built specifically for the era of AI coding agents. The problem is simple: when an AI writes code faster than you can read, you cannot just scroll through a standard git diff and type out file paths to ask for changes. lgtm fixes this by living right next to your agent in a terminal pane. As the AI writes, the view updates instantly. The coolest part is that you can use vim motions to jump straight to a line, add a note, and send it back to the agent without ever retyping a file name. It even tracks where comments go when the code shifts.

🆔 @hackernewsgithubprojects

📦 browser-use/jev-ultrafast

Jev Ultrafast: The Browser Agent That Thinks Faster

Jev Ultrafast is the browser agent that decides its next move without guessing. It scans your screen to see every button and text box, then picks exactly what to click or type in a single step. Instead of looking at pictures of your screen, which is slow and blurry, it reads the actual structure of the page directly. This means it understands where things are instantly. The project proves this speed by searching for a flight from Zurich to London in just seven seconds. It types the city names and clicks the dates all on its own. Because it skips the visual guesswork, it works much faster than traditional tools.

📰 https://news.ycombinator.com/item?id=49735979

🆔 @hackernewsgithubprojects

📦 starnodes2024/comfyui-yue2-trainer

ComfyUI-YuE2-Trainer: Train Your Own Music Style

You can now teach a massive AI model to sound exactly like your specific songs without needing a computer science degree. This tool adds training nodes directly into ComfyUI, allowing you to feed it your own music files and a simple trigger word. Instead of writing code, you just drag in your audio, let it learn the specific style, instrumentation, and vocal timbre of your tracks, and it generates a ready-to-use adapter. This solves the problem of needing complex command-line tools to customize AI music generation. It is particularly useful for artists who want consistent sound profiles for their projects.

🆔 @hackernewsgithubprojects

📦 eval-exec/neomacs

Neomacs: The GPU-Powered Emacs Rewrite

Neomacs is the text editor rewrite that finally lets Emacs run on your graphics card. For forty years, Emacs relied on aging code that could not keep up with modern screens, so this project rebuilt the entire engine in Rust to handle rendering directly on your GPU. This means switching between documents or scrolling through files now happens at your monitor’s full refresh rate, making the interface feel alive rather than static. The best part is that it still runs your existing configuration exactly as before, so you get the speed of a game engine without losing your setup.

📰 https://news.ycombinator.com/item?id=49743386

🆔 @hackernewsgithubprojects

📦 mykeura/minimalist-themes-for-hermes

Minimalist Themes for Hermes Desktop: 18 Warm Palettes

The Hermes Desktop interface just got eighteen new, warm light color options that stick with you even when you switch between different work profiles. Most theme plugins force you to pick a new look every time you change your active profile, which is annoying if you like a consistent, readable workspace. This project fixes that by keeping your chosen minimalist palette selected automatically while you move between profiles. The palettes are intentionally light and warm, avoiding the usual dark mode fatigue, and they only affect the desktop interface, leaving your command line tools untouched.

🆔 @hackernewsgithubprojects

📦 itsjazii/pane

Pane: Track All Your AI Subscriptions in One Tray

Pane lets you check your remaining AI quota without opening a single browser tab. If you juggle subscriptions for tools like Claude, Codex, and Cursor, you know the frustration of logging into separate dashboards just to see if you have enough credit left for your next task. This free, open-source Windows app solves that by living quietly in your system tray. It automatically finds the login tokens your coding tools already have, queries each vendor directly for your real usage limits, and projects whether you will run out of credits before your weekly reset.

🆔 @hackernewsgithubprojects

📦 kubetail-org/kstack

kstack: AI for K8s

kstack is the AI skill pack that finally makes Kubernetes troubleshooting feel like chatting with a senior engineer. Instead of you manually typing out complex commands to hunt down errors, it lets you just describe the problem in plain English, like why a specific app is crashing or how much memory a service is using. It then quietly does the heavy lifting in the background, pulling logs and metrics so you get clear, useful answers without drowning in technical jargon or burning through your AI credits. If you work with cloud infrastructure, this is the shortcut that keeps you in control while doing the tedious detective work for you.

📰 https://news.ycombinator.com/item?id=48045711

🆔 @hackernewsgithubprojects

📦 rizqinrr/viserys-agent

Viserys Agent: The Workflow for AI Coders

Viserys Agent is the workflow pack that stops your AI coding assistant from guessing. It forces your AI agent to follow a strict six-step process, from defining the problem to shipping the code, instead of just spitting out whatever it thinks you want. The coolest part is that it includes specific skills like doubt-driven development, which tells the agent to actively challenge its own assumptions before writing a single line of code. This turns a chatbot into a disciplined engineer that actually checks its work. If you are using AI to write code, this is the layer that keeps it from hallucinating its way into a mess.

🆔 @hackernewsgithubprojects