📦 nvidia/personaplex

Talk to Any Character Instantly with PersonaPlex

Make an AI voice that actually listens and reacts like a real person. PersonaPlex is a real-time speech model from Nvidia that lets you talk with any character you invent. Instead of just typing text, you can pick a specific voice and give the AI a detailed role, like a friendly teacher or a busy customer service agent. It is built on the Moshi architecture, which means it processes your words and speech at the same time, creating a natural back-and-forth flow. The coolest part is that it can handle interruptions and pauses, so it feels like a genuine conversation rather than a robotic exchange.

🆔 @hackernewsgithubprojects

📦 advaitpaliwal/feynman

Feynman: The AI Research Agent

Feynman is an open source AI agent that reads scientific papers, checks their code, and writes summaries you can trust. It solves the problem of information overload by acting as a tireless lab assistant. You ask it to review a topic, and it searches databases, compares claims against actual code repositories, and cites its sources. It even runs four distinct internal agents to research, review, write, and verify facts. This means you get a verified research brief instead of a generic summary. If you are a developer or scientist, this tool turns hours of manual reading into a few minutes of automated, accurate analysis.

🆔 @hackernewsgithubprojects

📦 uvwt/agentdock

AgentDock: ChatGPT Runs Your Computers

AgentDock lets you take full control of your physical computers and remote servers directly from a standard web browser chat. Instead of keeping your AI locked inside a secure sandbox, this tool hands the model safe, permission-bound access to your actual machine so it can write code, run commands, and deploy projects in the real environment. It’s like giving your favorite chat assistant a secure remote control for your digital workspace, letting you manage multiple devices from one conversation without juggling endless terminal tabs or burning through premium coding credits.

🆔 @hackernewsgithubprojects

📦 huggingface/funes

Funes: Give Your AI Agents a Long-Term Memory

Stop letting your coding assistants forget everything the moment you close the terminal. Funes solves this by building a durable, searchable memory of every past session you’ve had with agents like Claude Code or Codex. It’s like giving your AI tools a persistent brain that remembers the decisions, dead ends, and rationale from weeks ago. The coolest part? Your entire work history becomes a portable dataset you can publish to the Hugging Face Hub. Suddenly, your teammate or even another machine can recall exactly how you solved a problem before. It’s not just about saving text; it’s about sharing the actual thinking process behind your code.

🆔 @hackernewsgithubprojects

📦 vllm-project/agentic-api

Run OpenAI Codex on Your Own GPUs

Point OpenAI's coding agent at your own hardware and let it run without sending your code to the cloud. The project called Agentic API sits in front of your local model server, taking over the messy work of remembering conversation history and executing tools. Instead of your app tracking every message, you simply send one request, and the server handles the rest. It even speaks the exact language Codex uses, so you can drive that powerful agent using open-source models you control. It is a clean way to build self-hosted AI applications without losing the convenience of stateful conversations.

🆔 @hackernewsgithubprojects

📦 topoteretes/cognee-integrations

Cognee Integrations

Give your AI agents a real long-term memory. The Cognee Integrations repo gives popular agent frameworks, like CrewAI and LangGraph, a persistent knowledge layer. Instead of forgetting everything after a session ends, these agents remember past conversations and facts. The cool part is it works across different tools, so your agent doesn't start from zero every time. It builds a lasting brain for your apps, making them feel much smarter and more personal. If you build agents, this is the missing piece for true continuity. Grab it to make your AI actually remember who it is talking to.

🆔 @hackernewsgithubprojects

📦 agenticoding/pi-agenticoding

pi-agenticoding: Your AI Coding Crew Manager

pi-agenticoding is the workflow layer that lets you define AI coding crews once and reuse them forever. It solves the problem of constantly re-explaining which models should handle which tasks. You save a specific workflow, like a critical code review, and the system automatically assigns the right models to the right roles. It uses a shared notebook to keep important decisions safe while discarding the messy working notes. This means you get consistent, high-quality code checks without manually steering every single step. It turns repetitive AI management into a simple, repeatable command that actually works. Stop wasting tokens on instructions and start focusing on the code that matters.

🆔 @hackernewsgithubprojects

📦 sparkplug604/praxis

Praxis: The Memory Layer for AI

Praxis gives AI agents a way to inspect exactly where their answers come from. Most AI tools just guess based on whatever text is nearest to your question, but Praxis tracks the trust, age, and origin of every piece of knowledge it uses. This means if two sources disagree, the agent flags it as a warning instead of silently picking one. It turns scattered files, spreadsheets, and history into a clear, searchable memory that stays on your local machine. No cloud, no black box. Just a transparent layer that helps AI remember what matters and show you its work. It is not a hosted service.

🆔 @hackernewsgithubprojects

📦 nateherkai/snagtime

SnagTime

SnagTime is the self-hostable scheduling app that finally puts you in complete control of your booking infrastructure. Unlike standard third-party tools, this open-source project gives you the entire source code to run your own custom calendar system on your own terms. It packs essential features like Google Calendar synchronization, custom workspace branding, and test payments through Stripe, but keeps things entirely private. You can get a credential-free local demo running in just five minutes using a built-in SQLite configuration, making it incredibly easy to test out before deploying to a hardened production server. It is the perfect solution for anyone wanting a beautiful, independent scheduling portal.

🆔 @hackernewsgithubprojects

📦 anujatk1999/open-media

Open Media Shot Composer

Open Media is the browser-based storyboard planner that helps you block out cinematic 3D scenes before generating AI videos or images. Instead of wasting time guessing complex text prompts, you can visually pose realistic characters, arrange objects, and frame exact camera angles directly in your browser. It includes specialized settings like over-the-shoulder shots and camera elevation, plus a timeline to animate your setups. The coolest part is a local protocol server that lets an AI coding agent drive the entire scene creator for you. It is a completely local tool that runs entirely in your browser to bring your creative vision to life.

🆔 @hackernewsgithubprojects

📦 mattiaippoliti/ciele

Meet Ciele, Your Open Source AI Teammate

You can now spin up custom AI chat assistants that actually know your business without sending your data to third-party vendors. Ciele is an open-source platform that lets you build, test, and host smart AI teammates directly on your own hardware using a single terminal command. It crawls your websites, indexes your files, and handles customer questions using a clever router that only uses AI when it absolutely has to, falling back to your team when it gets stuck. It runs completely offline with any local model, giving you full control over your data.

🆔 @hackernewsgithubprojects

📦 jetbrains/benjamin-plus-skill

Benjamin Plus Skill

A single text injection cuts coding agent bills by nearly eighteen percent without touching quality. Benjamin Plus Skill is a set of rules that changes how agents look things up, not what they build. It teaches them to gather facts in one pass, read only what they need, and stop checking on finished jobs. The result is less wasted effort and lower costs for every task. If you run coding agents, this is the most efficient way to keep your budget from bleeding out. It is not a new tool, just a smarter way to use the ones you have.

🆔 @hackernewsgithubprojects

📦 swival/swival

Swival: Coding With Your Own Models

Swival lets you run a fully autonomous coding agent using your own local AI models. Most coding assistants force you to rent expensive cloud compute, but this tool works with the models you already have running on your own hardware. It handles the tricky part of keeping a smaller model focused so it does not forget its plan halfway through a task. You just point it at your model and give it a job, like fixing a bug or refactoring a file, and it takes care of the rest. It is a great way to keep your data private while still getting real coding help.

🆔 @hackernewsgithubprojects

📦 3641397194-wq/gpt6-astra

GPT6 Astra: The Six Seat AI Workbench

GPT6 Astra is the desktop workstation that turns six major language models into one unified command center. It solves the problem of juggling different AIs by giving each one a dedicated seat inside a single dark-themed application. The surprising feature is its ability to write custom instruction files directly into each model’s local configuration folder, effectively giving every AI a unique personality and set of rules tailored to your workflow. It handles the technical details of backing up original settings and verifying changes automatically. This gives you a persistent, customized setup for every model you use, all managed from one simple interface.

🆔 @hackernewsgithubprojects

📦 edvardgrishin27/nativeprompt

nativeprompt: The Prompt Translator You Didn't Know You Needed

Rewrite your AI prompts so they actually work on the specific model you are using right now. nativeprompt detects whether you are sitting in front of Claude Code or Codex and rewrites your text using that vendor's official rules. It fixes things like removing step-by-step instructions from reasoning models or dropping shouting caps that cause over-triggering. Every edit comes with a link to the official documentation, so you know exactly why it changed. The best part is that it runs completely offline on your existing subscription. There are no API keys, no extra bills, and no data leaving your machine.

🆔 @hackernewsgithubprojects

📦 sonicloudco/recorder

Recorder: The Tiny Card That Records Everything

Snap this tiny card onto any surface to capture high-fidelity audio without holding a bulky recorder. The project provides a complete toolkit for developers to bridge the gap between that physical hardware and your digital apps, handling everything from the initial Bluetooth handshake to complex file management. It solves the annoying problem of juggling multiple devices by giving you a single, unified interface to start, pause, and download recordings from phones or desktops. The real magic is in the seamless integration; it streams real-time audio data directly to your software, letting you process or transcribe conversations on the fly rather than waiting for a file transfer.

🆔 @hackernewsgithubprojects

📦 mazer666/phairplay

PhairPlay: Free AirPlay Receiver for Android TV

PhairPlay is the open-source Android app that turns your TV into a legitimate AirPlay receiver without needing an Apple TV. You can finally mirror your Mac or iPhone screen directly to any Android TV or Fire Stick, which is incredibly handy if you don't want to pay for proprietary hardware. The project handles the complicated encryption and video decoding in the background so you just see your desktop on the big screen. It is completely free, has zero ads, and works entirely on your local network, meaning your privacy stays intact.

📰 https://news.ycombinator.com/item?id=49599932

🆔 @hackernewsgithubprojects

📦 trailhq/graft

Graft: Give AI a Memory Map

Graft turns your entire codebase into a readable map that AI agents actually understand, cutting their work time by sixty percent. Usually, coding tools start every new task by blindly searching through files, wasting time and money. This tool solves that by building a folder of linked notes that explain exactly what every part of your system does in plain English. It is not a hidden database, but a local cache that stays fresh automatically. Since it uses standard files, any tool can read it instantly. You get faster results and better answers without changing your setup.

🆔 @hackernewsgithubprojects

📦 zorost/ai-engineering-lab

AI Engineering Lab

AI Engineering Lab is the self-paced course that treats you like a real engineer from day one. Instead of jumping between random tutorials, you build one continuous project for a fictional freight company over twenty-four weeks. That single story keeps your code connected, so your early data becomes the foundation for your final cloud deployment. It is free, fully open, and requires no signup. If you want to stop guessing what to learn next and start building systems that actually work, this is the map you need.

🆔 @hackernewsgithubprojects

📦 useagenthq/useagent

useagent: The AI Coworker

Your AI agent just got a full computer and a job description. useagent is an open source platform that turns coding assistants into actual teammates who can build real things like websites, spreadsheets, and code changes. Instead of just chatting, these agents work inside secure cloud machines, using your specific tools and company knowledge to finish tasks from start to finish. You can even run them right from Slack, just like pinging a human colleague. The coolest part is that every action is logged, so you can always see exactly what the agent did and why. It’s basically hiring a tireless intern who never sleeps and never forgets a step.

🆔 @hackernewsgithubprojects