📦 researai/autofigure

AutoFigure: AI-Powered Scientific Illustration Generation

AutoFigure is the research tool that turns complex scientific descriptions and academic papers into publication-ready visual diagrams. By leveraging a dual-agent system that performs iterative generation and evaluation, it transforms raw text into high-quality SVG or XML outputs compatible with common diagramming software. Users can extract methodology directly from PDFs or provide natural language prompts to create custom visuals. With integrated AI-powered enhancement features, it offers an efficient way to beautify your work while maintaining scientific rigor. Whether you are drafting a paper or building a presentation, this system simplifies the visual communication of technical concepts.

🆔 @hackernewsgithubprojects

📦 unconv-ai/un-0

Generating Images with Coupled Oscillators

Create images by integrating the phase dynamics of a population of coupled oscillators rather than relying on traditional diffusion or adversarial methods. This project explores how dynamical systems function as a specialized computing substrate, potentially offering a significantly more energy-efficient alternative to standard digital AI accelerators. The repository provides a reference implementation in PyTorch, complete with training pipelines for CIFAR-10 and ImageNet-64, alongside pretrained checkpoints that you can use to generate samples immediately. This approach demonstrates a unique, physics-inspired path for future machine learning hardware, showing that alternative computational models can produce high-quality results while rethinking standard AI architecture.

🆔 @hackernewsgithubprojects

📦 abundantbeing/hermes-browser-extension

Connect Your Browser to Hermes Agent

Hermes Browser Extension is the side panel tool that bridges your active web context to the Hermes Agent runtime. This browser-native extension communicates directly with your local or remote Hermes API server, allowing you to seamlessly integrate your models, tools, and memory into your daily workflow. It captures readable page text, metadata, and links, wrapping them as untrusted context before sending them to your configured sessions. With features like voice dictation and a sleek, themeable interface, this alpha tool lets you harness the power of your existing Hermes setup to analyze and interact with content right inside your browser.

🆔 @hackernewsgithubprojects

📦 asimons81/hermes-vault

Secure Your AI Agent Credentials with Hermes Vault

Hermes Vault functions as a local-first credential broker and encrypted storage system designed specifically for managing secrets used by AI agents. It effectively solves the problem of insecure credential handling by scanning for plaintext secrets, encrypting sensitive data in a local SQLite vault, and implementing time-bound access through lease management. The project provides an audit-ready environment where operators can verify credential health, execute policy-gated access, and use a local dashboard for visibility. By integrating with the Model Context Protocol, it allows agents to securely request temporary environment variables, ensuring that your sensitive tokens remain protected while maintaining operational agility.

🆔 @hackernewsgithubprojects

📦 abhi1693/openclaw-mission-control

OpenClaw Mission Control: AI Agent Orchestration Dashboard

OpenClaw Mission Control is the operations platform that provides unified governance for distributed AI agent workflows. It serves as a central dashboard where teams can manage organizations, assign tasks, and monitor multi-agent activity in one place. By incorporating explicit approval flows and gateway-aware orchestration, the system ensures that sensitive actions are verified and runtime environments remain connected across teams. With a robust API-first design, it allows operators to handle the full agent lifecycle and audit system actions for better accountability. This platform is perfect for teams needing a reliable, self-hosted interface to scale their AI operations efficiently.

🆔 @hackernewsgithubprojects

📦 tddworks/baguette

Control iOS Simulators Headlessly with Baguette

Manage and control iOS simulators directly from your terminal or a browser by bypassing standard GUI tools. This tool allows you to boot devices, stream screens at 60 frames per second, and inject precise inputs like taps, multi-finger gestures, and hardware button presses. You can even pipe your Mac webcam directly into simulator apps or inspect the accessibility tree for deeper insights. Whether you are running a single device or a massive simulated device farm for testing, it provides a powerful, headless interface to streamline your development workflow. Take full command of your iOS testing environment today.

📰 https://news.ycombinator.com/item?id=48690901

🆔 @hackernewsgithubprojects

📦 workweave/router

Slash AI Costs with Model Routing

This repository provides a model router designed specifically for agentic systems to optimize prompt delivery. It routes each prompt to the most suitable model in under fifty milliseconds, allowing developers to reduce operational costs by forty to seventy percent simply by changing an endpoint. The project includes a robust dashboard for monitoring statistics, tools for managing model configurations, and a proxy service that handles intelligent routing decisions. By dynamically matching requests to the appropriate AI model, this tool helps maintain performance while significantly lowering expenses. Take control of your agentic workflows and start driving down your infrastructure costs today.

📰 https://news.ycombinator.com/item?id=48688700

🆔 @hackernewsgithubprojects

📦 avivk5498/the-claude-protocol

Stop Losing Context: The Claude Protocol Explained

Orchestrate multi-agent workflows within Claude Code by implementing an enforcement layer that tracks every task and decision. This project uses the Beads CLI to isolate work in dedicated git worktrees, ensuring that every unit of work remains traceable from creation to merge. It prevents common issues like lost session context or unsupervised code edits by utilizing thirteen automated hooks that validate actions and enforce strict development discipline. By automating knowledge capture and maintaining persistent memory across sessions, it allows specialized agents to execute fixes reliably, providing a structured, professional workflow for complex coding projects.

🆔 @hackernewsgithubprojects

📦 secondly-com/openphone

OpenPhone: The AI-Native Android OS

Transform your mobile device into a truly agentic system with this AI-native operating system built on Android. Instead of a standard chatbot, it integrates a privileged system-level agent that can actively observe your screen, operate applications, and monitor phone events in real time. The platform excels at proactive work by handling background tasks, remembering your commitments, and allowing for detailed auditability of every action taken. It features a robust runtime for managing automated workflows, giving you a powerful, transparent, and intelligent way to interact with your phone's capabilities and apps. Explore this project to redefine your smartphone experience today.

📰 https://news.ycombinator.com/item?id=48678160

🆔 @hackernewsgithubprojects

📦 mnapoli/vibephp

VibePHP: A Runtime That Hallucinates Your Web Server

VibePHP is a experimental PHP engine that executes code entirely through AI hallucination rather than traditional compilation or interpretation. When a request arrives, the engine sends your source code to an AI model that reads the logic and spontaneously invents the missing data, including database records, API responses, and clock values, to produce a believable HTTP result. Built on Laravel and the AI SDK, it effectively turns any PHP file into a functional, ever-changing web page without requiring a real backend infrastructure. This project challenges standard execution models by prioritizing creative improvisation over technical correctness or performance.

📰 https://news.ycombinator.com/item?id=48690737

🆔 @hackernewsgithubprojects

📦 saihgupr/androidtvbluetooth

Control Your Android TV Bluetooth Devices with ADB

AndroidTVBluetooth is the lightweight utility that finally gives you one-click control over your television Bluetooth connections. This app provides a simple interface to connect or disconnect paired devices instantly, while also offering robust support for automation through terminal commands. You can trigger specific device connections by name or MAC address using broadcast intents, making it perfect for custom remote mapping or smart home scenarios. By providing direct ADB shell access to Bluetooth profiles, this project removes the friction of navigating system settings. Take full control of your entertainment setup today with this powerful, open-source automation tool.

🆔 @hackernewsgithubprojects

📦 uellenberg/insert

Build Self-Modifying Code with Insert

Create programs that rewrite their own source code using Insert, a specialized programming language designed for building self-modifying applications. By leveraging unique features like markers and bindings, you can track specific locations within your code to update logic or data dynamically at runtime. This allows for fascinating implementations, such as a pong game where each new frame is rendered by modifying the source code of the next generation. The compiler, written in Rust, translates your logic into C while preserving the structure needed for these persistent quine-like behaviors, offering a powerful tool for exploring creative code evolution.

🆔 @hackernewsgithubprojects

📦 light-and-ray/minimalistic-comfy-wrapper-webui

Simplify Your ComfyUI Experience with MCWW

Minimalistic Comfy Wrapper WebUI is the inference focused interface that finally simplifies your complex ComfyUI workflows into a clean mobile friendly experience. This tool hides the complex node web and gives you a stable environment where your UI states are saved locally in your browser. You can easily manage queues with priority control, utilize convenient prompt presets for one click generation, and even perform basic image editing directly in your browser. Since it dynamically adapts to your existing workflows, you simply title your nodes to define your inputs and outputs, making it a perfect, lightweight alternative for daily generation.

🆔 @hackernewsgithubprojects

📦 nolangz/pixel2motion

Turn Your Static Logo Into a Smooth Animation

Transform static raster logos into professional motion graphics using a systematic workflow that bridges the gap between design and development. This project provides an open-source skill that guides you through converting PNG or JPG files into clean, motion-ready SVG geometry. It focuses on structural integrity, ensuring each part of your logo is optimized for smooth, choreographed animation using established design principles. By generating interactive HTML demos and detailed motion previews, it allows for precise quality assurance testing, helping you create refined brand reveals or loading states that stay faithful to your original design. Start animating your brand with clarity and control.

🆔 @hackernewsgithubprojects

📦 matiasbattocchia/open-bsp-ui

Build Your Own WhatsApp & Instagram Interface

Manage your messaging workflows with a self-hostable interface designed for WhatsApp and Instagram. This open-source platform lets you handle real-time conversations, organize contacts, and build custom message templates with ease. You can also configure AI agents to automate responses, track usage statistics, and manage team roles across multiple organizations through a clean, mobile-first design. Built with React and Vite, the interface offers a responsive experience that keeps your communication centralized and efficient. Use this tool to maintain full control over your customer interactions while simplifying how you connect with your audience across different social channels.

🆔 @hackernewsgithubprojects

📦 gbsoss/skill-from-masters

Build AI Skills Using Proven Expert Methodologies

Skill From Masters is a tool designed to help you create AI skills based on the verified methodologies of domain experts. Instead of guessing how to structure your agent tasks, this system searches a curated local database and the web to surface established frameworks, common pitfalls, and golden examples from respected figures across fields like product management, leadership, and engineering. By cross-validating these sources before any code is generated, it ensures your skills embody professional-grade expertise. It integrates directly into your workflow to bridge the gap between abstract knowledge and actionable, high-quality automation. Build smarter by learning from the best.

🆔 @hackernewsgithubprojects

📦 nityeshaga/claude-code-essentials

Optimize Your Workflow with Claude Code Essentials

Enhance your development routine by integrating a collection of specialized plugins designed to streamline coding, architecture, and deployment tasks. This project provides a powerful marketplace for tools that improve productivity, including advanced command-line utilities for code reviews, gem upgrades, and documentation generation. You can adopt a brutally honest DHH-style Rails reviewer, implement safety hooks to prevent accidental git pushes, and apply algorithmic workflows to simplify requirements and cut unnecessary technical complexity. Whether you need expert deployment guidance or refined prompt engineering skills, these resources help developers build cleaner, more efficient applications while maintaining high standards for code quality.

🆔 @hackernewsgithubprojects

📦 vila-lab/opai-bench

Is AI Text Detection Actually Reliable? New Research Says No

AI text detection is surprisingly unreliable when humans and machines collaborate on writing. OpAI-Bench is a comprehensive benchmark that tracks how AI editing impacts detection accuracy across nine distinct revision stages. Unlike standard tools that evaluate only static drafts, this project monitors text as it shifts through polishing, paraphrasing, and expanding. Researchers found that mixed-authorship intermediate versions are often the most difficult for detectors to identify correctly. This repository provides a complete pipeline to evaluate various detection models, helping developers understand how authorship signals evolve during the writing process and ultimately improving the design of future detection systems.

🆔 @hackernewsgithubprojects

📦 l1ad/claude-task-viewer

Visualize Your AI Workflow with Claude Task Viewer

Claude Code tasks are now viewable through a dedicated web-based Kanban dashboard. This tool monitors your local task files to provide a real-time, visual interface for everything the AI is working on, solving the problem of terminal-only visibility. It displays live activity across multiple sessions, tracks task dependencies, and features a Gantt-style timeline for duration analysis. You can receive desktop notifications when tasks complete and manage your project history with automatic session archiving. By watching your filesystem, it creates a persistent bridge into your AI automation, keeping your workflow organized and accessible at all times.

🆔 @hackernewsgithubprojects

📦 amanaiproduct/personal-os

PersonalOS: Turn Notes Into Actionable AI Tasks

PersonalOS is the task management framework that converts your chaotic brain dumps into organized, goal-driven actions. By leveraging local AI, it allows you to capture unstructured thoughts in a backlog and transform them into prioritized tasks linked directly to your defined objectives. It uses smart deduplication to keep your workspace clean and integrates with AI assistants via the Model Context Protocol to help you focus on what actually matters. Whether you need to manage P0 priorities or reference your knowledge base, this system keeps your data local, organized, and ready for your daily workflow.

🆔 @hackernewsgithubprojects

📦 0xshug0/audio.cpp

Run Modern Audio Models Locally with C++

audio.cpp is the high-performance inference engine that brings modern audio models to native C plus plus. By leveraging the ggml library, this project enables you to run tasks like text-to-speech, voice conversion, and speech recognition locally without any Python dependencies. It is specifically optimized for CUDA, delivering significantly lower latency and faster throughput than traditional reference paths. Whether you need VAD, diarization, or complex multi-step pipelines, this framework provides reusable, efficient building blocks for production-style audio tasks. Experience faster, more portable audio processing and take full control over your local machine learning workflows today.

🆔 @hackernewsgithubprojects