📦 theagentplane/tokenops

TokenOps

TokenOps can cut your AI agent’s wasted spending by sixty-five percent just by checking the bill before every single step. It treats an entire multi-agent workflow as one single job with one shared budget, not a bunch of separate calls. Instead of letting a complex task quietly burn through your credits, it watches the total spend in real time and stops the run the moment it hits the limit. This prevents that sneaky scenario where dozens of tiny, cheap actions add up to a surprisingly huge final invoice. It is basically a guardrail for your AI’s money.

📰 https://news.ycombinator.com/item?id=49574349

🆔 @hackernewsgithubprojects

📦 shiphrahx/ai-for-engineering-leaders

AI for Engineering Leaders: Your New Copilot for Management

There is a specific, practical playbook on GitHub called AI for Engineering Leaders that turns the chaotic art of managing teams into a structured, copy-and-paste workflow. It isn't just a vague list of tips; it is a library of over one hundred ready-made prompts and skills designed for the messy reality of engineering leadership. If you are a manager, director, or VP, this repo hands you the exact words to use for your next executive status update, a tricky performance review, or a critical incident postmortem.

🆔 @hackernewsgithubprojects

📦 johndpope/llama-cpp-turboquant

llama-cpp-turboquant

llama-cpp-turboquant is the local inference engine that squeezes massive language models onto your regular computer hardware. Most people think running big A.I. models requires expensive cloud servers, but this project lets you run them directly on your laptop or desktop without sending your data anywhere. It uses clever math called quantization to shrink the model files, which means you get faster speeds and less memory usage. If you care about privacy or want to play with A.I. offline, this is the tool that makes it possible right now. Stop renting your A.I. and start owning it. Your next project runs locally, and it is actually quite impressive.

🆔 @hackernewsgithubprojects

📦 worldflowai/everything-claude-code

How to Make Claude Code Actually Work

An Anthropic hackathon winner just open-sourced the exact configuration files he used to build real products with AI. This is not a new coding tool; it is a complete toolkit of specialized agents and rules that you plug directly into your existing setup. The coolest part is how it handles memory. Instead of forgetting everything when a conversation gets long, it automatically saves your progress and learns patterns from your past sessions so it gets smarter every time you work. It turns a generic chatbot into a disciplined senior developer that follows your specific coding standards.

🆔 @hackernewsgithubprojects

📦 code-yeongyu/lazycodex

LazyCodex: The Missing Memory for AI Coders

LazyCodex solves the most annoying problem in AI coding: your AI assistant forgets everything the moment you close the window. It acts as a persistent brain for OpenAI Codex, remembering your project structure, planning complex tasks, and actually verifying that the code works before it tells you it is done. You do not just get answers; you get a disciplined workflow that treats your codebase like a real software project. It installs in one line, needs no global setup, and works right inside your existing Codex environment. If you are tired of re-explaining your project to a chatbot every single time, this is the fix.

🆔 @hackernewsgithubprojects

📦 jakubantalik/libraries.dev

Libraries.dev: The Glow Engine

Libraries.dev builds visual effects for artificial intelligence interfaces that actually feel alive. Instead of a static spinner, these libraries give a thinking machine a physical presence, using animated beams and flowing liquids to show you exactly what the software is doing right now. It solves the problem of waiting for complex calculations without knowing if your computer is stuck or still working. The most interesting part is the thinking orbs, which mimic human thought patterns, making the machine feel responsive rather than frozen. By turning invisible data processing into visible motion, it transforms a dry technical wait into an engaging experience.

🆔 @hackernewsgithubprojects

📦 coulsontl/ai-toolbox

AI Toolbox: Control Every AI Coder in One App

Stop juggling a dozen different config files and scattered API keys for every AI assistant you own. AI Toolbox is a dedicated desktop app that centralizes the settings for major coding tools like Claude Code, OpenAI Codex, and Gemini CLI into one clean, visual dashboard. It handles the messy details, such as switching providers, managing prompts, and syncing your setup to remote servers or WSL environments, so you never have to hunt through hidden folders again. The real magic is the local proxy gateway, which lets you swap models or track usage without changing a single line of code.

🆔 @hackernewsgithubprojects

📦 tripolskypetr/backtest-kit

Backtest Kit: Live Trading on a Single Engine

Backtest Kit is the trading engine that finally lets you test and run the exact same strategy without rewriting a line of code. Most tools force you to build a simulation, then build a separate live bot, and then spend your weekends patching the two so they actually agree. Here, you write your logic once, and the system handles both the historical replay and the real-time market execution. The coolest part is that it uses the same underlying engine for both modes, which means the results you see in your backtest are literally the same code path that trades your money in live.

🆔 @hackernewsgithubprojects

📦 deeplethe/lethe

Lethe: The AI Memory That Forgets

Lethe is the AI memory system that finally learns to forget. While most tools just keep piling up facts until they get messy, Lethe treats every piece of information like a leaf floating in water, letting old or wrong details naturally sink and disappear. The coolest part is that it can cryptographically prove a fact is gone, which is huge for privacy. It runs entirely on your computer with no cloud calls. If you are building agents, this is how you keep them clean and accurate without the data bloat.

🆔 @hackernewsgithubprojects

📦 kronop/vibe-aso

Vibe ASO: The AI Tool That Localizes Your App for 50 Countries

A single AI assistant can now handle the entire App Store optimization process for your mobile app, handling everything from keyword research to translating your interface into fifty different languages. It solves the massive, repetitive problem of manually managing metadata, screenshots, and pricing for every single market you want to enter. By using a specialized tool, you can verify that your app looks professional and speaks the local language correctly before it ever reaches the App Store. You get a polished, global-ready product without hiring a team of translators or spending weeks on manual updates.

🆔 @hackernewsgithubprojects

📦 wangpeterxf/session-harbor

SessionHarbor: Verified Backup for AI Sessions

SessionHarbor takes every conversation you have with your AI coding assistant and creates a verified, tamper-proof copy on an external drive. It solves a real problem: when you delete a chat to free up space, you usually lose that context forever, but here, the backup is cryptographically checked before any local file is allowed to go. This means you can safely reclaim disk space on your Mac or Windows machine, knowing you can restore the exact same data later without guesswork.

🆔 @hackernewsgithubprojects

📦 mengto/sublevel-studio

Sublevel Studio: The 3D Portfolio That Is Actually A Game

Sublevel Studio is the web portfolio that drops you into a playable, three-dimensional Kyoto lobby instead of a boring landing page. Forget the standard hero section; this project wraps a creative studio's work in an immersive, interactive room where you can actually touch the environment. It features a working arcade cabinet, a basketball hoop, and old-school C.R.T. monitors that lead you into the case studies. The coolest part? The whole thing is crammed into a single file with no build steps or servers required. You can run it locally in seconds.

🆔 @hackernewsgithubprojects

📦 adkid-zephyr/anti-defensive-writing-skill

Anti-Defensive Writing: The Prompt Pack

Anti-Defensive Writing is the academic writing tool that stops your paper from accidentally attacking itself. It solves the problem where AI assistants turn strong research into timid, self-censoring reports full of apologies and weak language. The core idea is simple: treat your paper like a press conference, not a project summary. You only present your strongest advantage and ignore the metrics where you lose. This lightweight skill pack includes copy-paste prompts that teach any AI to enforce this principle. Instead of listing every failed attempt, it restructures your narrative to highlight exactly why your work matters. It ensures every experiment serves a specific argument, not just filling space.

🆔 @hackernewsgithubprojects

📦 wnby/photo-relic-editorial

Photo Relic Editorial: Photos Turned into Paper Memories

Photo Relic Editorial is the creative skill that turns real photos into quiet paper memories. It takes a genuine photograph, keeps the top half perfectly truthful, and compresses the bottom into a minimal, ink-style print that looks like the image was pressed into warm paper. It is not a generic filter; it extracts the actual light, structure, and mood of your shot to create a restrained, editorial artwork. The result feels less like digital editing and more like a tangible, hand-pressed relic that captures the emotional weight of the moment. It is a clever way to transform everyday snapshots into a cohesive, artistic series.

🆔 @hackernewsgithubprojects

📦 vectorspacelab/arex-skill

AREX-Skill: Teach AI to Code Like a Researcher

Turn working code into ready-to-use instructions for artificial intelligence. This project scans popular software libraries and extracts specific, tested capabilities into standalone packages that coding agents can execute directly. Instead of guessing how to run an experiment, the agent loads the exact steps needed for the task at hand. The result is a massive library of verified skills that help autonomous systems complete complex research goals without getting stuck. It transforms scattered documentation into actionable power. If you build tools with artificial intelligence, this is how you give them real muscle.

🆔 @hackernewsgithubprojects

📦 qwenlm/qwen-drive-1.0

Qwen-Drive: The AI That Thinks Before It Drives

Qwen-Drive turns raw camera feeds into safe driving decisions by letting a language model explain its reasoning before moving the car. Built by the Qwen team, this single model handles everything from spotting obstacles to predicting the next five seconds of motion. The real surprise is that it doesn't just guess a path; it generates a logical rationale for why that path is safe, making its thinking visible and checkable. This is a major step toward self-driving cars that can actually be understood by humans. If you are into AI, this is how explainability meets autonomous driving.

📰 https://news.ycombinator.com/item?id=49577104

🆔 @hackernewsgithubprojects

📦 pis10/tracesurface

TraceSurface: Find Hidden APIs in Frontend Code

TraceSurface takes a single login page and maps out every hidden backend connection buried in the JavaScript. It launches a real browser to catch live requests, then pairs that with a static code analysis to find endpoints you never clicked. The magic is in the replay: it strips your login credentials and sends the requests back to see who gets through without permission. One scan turned a simple captcha call into over a hundred exposed endpoints. It is a powerful way to audit your frontend for security gaps before attackers find them. If you ship web apps, this tool is a must-have for finding weak links in your API.

🆔 @hackernewsgithubprojects

📦 miaai-lab/qwen3.8-flash-next-single-dgx-spark

One Card, Giant Model

Run a massive, multimodal AI model on a single workstation GPU. This project lets you serve a giant vision-language model that handles text, images, and video using only one piece of hardware. The real magic is the memory trick. Instead of forcing the entire model into the graphics card, it parks the huge lookup tables in system memory and fetches them on demand. This simple trick lets a 99-gigabyte model fit inside a device with just 121 gigabytes of unified memory. It turns a modest desktop chip into a capable server for serious AI tasks without buying a rack of data center gear.

🆔 @hackernewsgithubprojects

📦 cloudgeni-ai/opengeni

OpenGeni: Run AI Agents on Your Own Hardware

OpenGeni is the self-hostable runtime that finally lets you run AI agents on your own hardware without losing control. It solves the trust gap in enterprise AI by keeping your data, credentials, and audit logs entirely inside your own infrastructure. The standout feature is the connected machine model, where you enroll your laptop or server as a first-class compute target. Instead of sending code to a vendor's cloud, the agent dials out to your machine to work directly with your files and git repos. It’s a durable, replayable session API that treats your hardware as an equal to managed sandboxes.

🆔 @hackernewsgithubprojects

📦 hwdsl2/docker-whisper

Docker Whisper: Your Private AI Transcriber

Docker Whisper is the self-hosted server that finally lets you process audio without sending it to the cloud. It runs OpenAI’s powerful speech recognition model directly on your own hardware, turning any voice recording into accurate text. The coolest part is speaker diarization, which automatically identifies exactly who is talking in a group conversation. This means your meeting notes know which person said what, all while keeping your data completely private. It even speaks the same API language as OpenAI, so you can switch your existing apps over with a single line change.

📰 https://news.ycombinator.com/item?id=47732384

🆔 @hackernewsgithubprojects

📦 yejy53/editable-design

Editable-Design: AI Design That Never Flattens

Editable-Design is the visual design tool that refuses to flatten your work. Instead of giving you a single, static image where changing one word breaks the whole picture, it builds your design using real, separate text layers and independent images. This means you can select any element and tweak it exactly as you need. It uses AI to handle the tricky parts of layout and style, but the result is fully editable code. So you get professional-level visual polish without being locked into a single, unchangeable screenshot. It is a smarter way to create graphics that actually work for your specific needs.

🆔 @hackernewsgithubprojects