AutoRAG: a self-evolving librarian agent that curates document search results into actionable knowledge units
A self-evolving librarian agent that searches document collections and curates results into direct answers instead of raw file paths.
Git --end-of-options: Preventing Argument Injection in Version Control
The --end-of-options flag in Git solves a critical security gap where untrusted input starting with a dash can be parsed as a command-line option, leading to argument injection vulnerabilities.
Roblox Android Remote Attestation and GrapheneOS Support
Roblox has implemented Android remote attestation to prevent cheating and modified apps, while explicitly announcing support for GrapheneOS with a locked bootloader.
CodeAlmanac: AI-Maintained Codebase Wiki for Coding Agents
CodeAlmanac is a local-first codebase wiki that uses AI agents to capture architectural decisions, invariants, and workflows in Markdown files stored directly in the repository.
claude-mem: a persistent memory system for Claude Code that preserves project context across sessions
A persistent memory compression system for Claude Code that preserves project context across sessions using semantic summaries and hybrid search.
ReadKinetic: A Local-First RSVP Speed Reader for Digital Books
ReadKinetic is a free, local-first speed reading application that uses Rapid Serial Visual Presentation (RSVP) to display words one at a time, reducing eye movement and increasing reading speed.
EU Court Rules VPNs Lawful for Bypassing Geo‑Blocks in Copyright Case
The Court of Justice of the European Union ruled that using a VPN to bypass a publisher's geographic block does not violate copyright law, affirming VPNs as lawful technical tools.
Imagin Raw: A Lightweight Open-Source Alternative to Adobe Bridge for Mac
Imagin Raw is a 9MB open-source macOS application designed for browsing, culling, and organizing RAW photos without the overhead of a managed library.
HN Hall of Fame: A Curated Archive of 3,100 Legendary Hacker News Links
The HN Hall of Fame is a curated directory featuring 3,100 of the most influential and high-scoring links from Hacker News history.
Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber Release
Google has released Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, focusing on token efficiency, low latency, and specialized cybersecurity capabilities for agentic workflows.
AI & Frontier Tech Roundup: Agentic Swarms, Robotics, and the Rise of Open-Weight Models
The frontier of AI is shifting toward autonomous agentic swarms, high-performance open-weight models like Kimi K3, and the integration of embodied intelligence into humanoid robotics.
hyperresearch: a deep research harness for Claude Code featuring a 16-step adversarial pipeline and a persistent source vault
A deep research harness for Claude Code that uses a 16-step adversarial pipeline and a persistent, searchable vault to produce high-fidelity research reports.
AI × Crypto Roundup: The Rise of the Agentic Economy
The intersection of AI and blockchain is shifting from simple chatbots to a complex economy of autonomous agents requiring specialized infrastructure for payments, identity, and verifiable compute.
Computable GPU Marketplace: Weekly GPU Capacity Trading and Auctions
Computable is a GPU marketplace that allows users to buy, sell, and redeem GPU hours across specific calendar weeks with instant liquidity and wholesale pricing.
ruler: a centralized instruction manager that synchronizes AI coding rules across multiple AI agents
A tool that centralizes AI coding assistant instructions into a single source of truth and automatically distributes them to the configuration files of various supported AI agents.
OpenAim: A Science-Based FPS Aim Trainer
OpenAim is a browser-based FPS aim trainer that utilizes a shared player model to personalize training drills and provide data-driven sensitivity tuning.
agent-service-toolkit: a full-stack template for deploying LangGraph agents with a FastAPI backend and Streamlit UI
A full-stack toolkit for building AI agent services using LangGraph, FastAPI, and Streamlit, providing a complete template from agent logic to user interface.
Jacobian conjecture counterexample in three dimensions explained
Terence Tao’s 2026 blog post presents an explicit polynomial map F: C^3 → C^3 with constant non‑zero Jacobian that is not invertible, providing a counterexample to the Jacobian conjecture in three dimensions.
potpie: a context graph that transforms codebases and development lifecycles into actionable knowledge for AI agents
Potpie creates a living context graph from your codebase and development tools, providing AI agents with the deep project knowledge required to plan, debug, and write code.
text-to-cad: what it is, what problem it solves & why it's gaining traction
A library of agent skills that enables AI to generate, inspect, and prepare CAD models and robot description files for fabrication and simulation.
agents: a framework for building real-time programmable multi-modal voice agents with integrated job scheduling
A framework for building real-time, programmable multi-modal AI agents that can see, hear, and speak, featuring integrated job scheduling and telephony support.
Apple Defeats Liability for iCloud CSAM Scanning in Amy v. Apple
A US court dismissed a lawsuit against Apple, ruling that Section 230 provides immunity for the company's decision not to scan private iCloud storage for Child Sexual Abuse Material (CSAM), prioritizing end-to-end encryption over proactive monitoring.
cube: an open-source semantic layer for defining governed metrics and dimensions across BI tools and AI agents
An open-source semantic layer that allows you to define metrics and dimensions once in code and expose them via APIs to BI tools and AI agents.
Federal Reserve Bank of Minneapolis Introduces Homeowners-to-Population Ratio (HPOP)
The Federal Reserve Bank of Minneapolis introduces the homeowners-to-population ratio (HPOP), showing U.S. homeownership at 53% versus the traditional 65% owner-occupancy rate.
agents: a multi-harness marketplace of agentic plugins and domain-expert agents for AI coding assistants
A marketplace of production-ready agentic building blocks that provides a single source of truth for plugins, agents, and skills across multiple AI coding assistants.
novu: Open‑source communication infrastructure that lets products and AI agents talk to users on any channel
Novu is an open‑source platform that offers a single API and unified conversation model for sending and receiving messages across inbox, email, SMS, push, and chat channels. It serves both product notification needs and AI‑agent communication (ACI), handling provider integrations, workflows, digests, and embeddable UI components.
Rediscovery of a Thriving Coral Reef off the Coast of Benin
Scientists have rediscovered a healthy coral reef ecosystem off the coast of Benin, West Africa, which had been presumed dead for sixty years.
Jane Street Incremental: A Library for Incremental Computations
Incremental is an OCaml library developed by Jane Street that optimizes computation graphs by only re-evaluating nodes affected by source data changes.
Poolside’s Model Factory, Laguna S, and Open Models: Eiso Kant on Building Code Models for AGI
Eiso Kant explains how Poolside’s Model Factory enables rapid eight‑week model cycles, why Laguna S relies on persistence and verification over raw intelligence, and why he favors many open foundation model companies over a few closed ones.
Stanford Robotics Seminar: Embodied Intelligence with Morphing Matter
Assistant Professor Lining Yao discusses leveraging morphing materials and mechanisms to create 'embodied intelligence'—hardware systems capable of programmability and decision-making through shape-changing and tunable materials.
Castle Map: An Open-Data Atlas of 2,435 Global Fortifications
Castle Map is a free interactive atlas featuring 2,435 curated castles, fortresses, and palaces across 129 countries, built using open data from Wikidata and Wikimedia Commons.
Promoting Agency in Human-AI Interaction: Stanford CS547 HCI Seminar
Matthew Jörke argues that AI systems should shift from task-automating assistants to augmentative advisors that preserve user agency through non-prescriptive support and the elicitation of qualitative context.
Drawing the Mona Lisa with GPT-5.6 Sol, Claude Fable 5, Grok 4.5, Gemini 3.6 Flash – Results and Analysis
In a colored‑pencil drawing arena, GPT-5.6 Sol produced the highest‑quality Mona Lisa and Starry Night reproductions while using far fewer tokens and lower cost than Claude Fable 5, Grok 4.5, and Gemini 3.6 Flash, which showed higher token usage, higher cost, or lower final similarity scores.
FreeInk: An Open Ecosystem for E-Readers
FreeInk is an open-source collective building a hardware-independent SDK, open firmware, and open-hardware e-reader core (de-link) to eliminate proprietary lock-in for e-paper devices.
Flock Safety ALPR Controversy: Patterns of Misrepresentation and Privacy Concerns
The ACLU has documented a pattern of Flock Safety providing false information to city councils and the public regarding its automatic license plate reader (ALPR) capabilities, data sharing with federal agencies, and partnerships with civil liberties organizations.
vLLM AFD Plugin: Disaggregating Attention and FFN for MoE Serving
The vLLM AFD Plugin introduces Attention-FFN Disaggregation (AFD), allowing Attention and FFN paths in Mixture-of-Experts (MoE) models to scale and execute independently to optimize serving throughput and latency.
Serving GLM-5.2 on NVIDIA B300 GPUs with vLLM
vLLM achieves production SLA compliance for GLM-5.2-NVFP4 on 24 NVIDIA B300 GPUs by implementing P/D disaggregation, speculative padding, and IndexerCache optimizations.
Hugging Face Integrates Nunchaku 4-bit Diffusion Inference into Diffusers
Hugging Face has integrated Nunchaku Lite into the Diffusers library, enabling 4-bit weight and activation (W4A4) quantization to reduce VRAM usage by up to 50% and improve inference speed by approximately 30%.
Health in ChatGPT Launch
OpenAI has launched Health in ChatGPT for U.S. users, allowing the AI to securely connect to Apple Health and medical records to provide personalized, context-aware health insights.
Online Speech Chilled by Surveillance: Insights from Hacker News Discussion
Many Hacker News commenters say they now speak less freely online because of state, corporate, and AI-driven surveillance, while others say they feel a duty to speak up or see no change.
ciso-assistant-community: a centralized cybersecurity management hub for governance, risk, and compliance
CISO Assistant is an open-source cybersecurity management platform designed to simplify Governance, Risk, and Compliance (GRC) through centralized data and framework decoupling.
Jellyfin Leadership Transition: Founder and Core Team Members Depart
Jellyfin project leader Joshua Boniface and core member Anthony Lavado have resigned alongside founder Andrew Rabert, citing burnout and personal priorities, while ensuring an amicable hand-off to the remaining team.
tracecat: an agentic security automation platform combining AI agents, low-code workflows, and case management
An agentic security automation platform that combines AI agents, low-code workflows, and case management to automate security operations.
osaurus: a native macOS AI harness with local agents, secure sandboxing, and on-device privacy filtering
A native macOS AI harness that provides local agents, memory, and secure code execution sandboxes, allowing users to own their AI context and identity offline.
mistral.rs: a high-performance, zero-config LLM inference engine with native agentic runtime and multimodal support
A fast, flexible LLM inference engine that provides zero-config execution for Hugging Face models with native support for multimodality and agentic tools.
opencodex: a universal provider proxy that enables any LLM to work with OpenAI Codex and Claude Code
A local proxy that translates Codex's API to allow any LLM provider (Anthropic, Gemini, Ollama, etc.) to be used within OpenAI Codex and Claude Code.
WordInWeb: Browser-Based DOCX Editor Aiming for MS Word Parity
WordInWeb is an open-source MIT-licensed browser-based DOCX editor that aims for near-MS-Word parity and runs entirely locally, as shown by user feedback on Hacker News.
AI and the Shift in Programming Difficulty
AI has shifted the primary challenge of software development from the recall of syntax and implementation to the critical judgment of architectural correctness and system integrity.
Claude Is Not a Compiler: The Rise of Vibe-Engineering
Bryan Mikaelian argues that LLMs like Claude are not mere compilers of natural language to code, but vertically integrated resources that enable 'vibe-engineering' by working across strategy, architecture, and implementation layers.
Immersive 3D Reconstruction of Grace Cathedral via Gaussian Splatting
Vincent Woo has created a high-fidelity digital twin of San Francisco's Grace Cathedral using 3D Gaussian Splatting and PlayCanvas, demonstrating the technology's ability to preserve historic architecture with extreme detail.